43 Visions for Complexity

The following sections are included:

• PREFACE

On a summer day in 2005 Jan Vasbinder visited me at the European Science Foundation. Jan had a dream: to set up a European institute dedicated to sciences for our future and based on the principles of the Santa Fe Institute. It was to be an independent institute that would grow a new kind of scientific research community, one emphasizing multi-disciplinary collaboration in pursuit of understanding the common themes that arise in natural, artificial, and social systems. It was a wonderful dream, and I bought into it right away. So I became founding father and initial funder of Institute Para Limes IPL. In March 2006 ESF hosted the first meeting of the founding fathers, in which the structure, governance and the first research agenda were decided upon…

We now have had 400 years of extremely successful reductionist science in which systems have been understood at finer and finer levels. We have also had 40 or so years of looking in the other direction, at how system behavior emerges from the interactions of its lower-level elements. This new movement in science, complexity, is not old and we stand very much at the beginnings of what it will bring…

Complex systems in their environments are open, emerging and dying, individuated and heterogeneous, multilevel organized and controlling from inside their boundary conditions. The phenomenological and theoretical reconstruction of their multi-level dynamics is the main epistemological challenge of complex systems science…

As society becomes more and more interconnected (we just started to add our environment to the net with the Internet of Things revolution), the net will be the place where most inter actions will take place…

The basic goal of the sciences is to point to, and explain, emergent phenomena: what we would not have guessed given what we knew before. This lack of predictability can come from a change of scale (more is different; physics), a change of descriptive language (lost in translation; the human sciences), or just patience on the part of the observer (self-organization; biology). Nothing worth knowing can be predicted…

Complexity science and big data science are challenging areas that require competences and inputs from different fields. These areas are interdisciplinary by their nature and problems and challenges that they are addressing are intrinsically related to the continuous evolution of our society…

From my personal point of view, complexity is the science of integration: integrating different mechanisms, scales, and dynamical rules to generate a new and emergent system. It integrates other traditional scientific disciplines in a common framework. Complexity science makes use of powerful tools of statistical physics, mathematics, and computer science, to bring back answers and even predictions to the originating disciplines. In particular, networks provide the topological skeleton through which interactions take place…

Pierre Teilhard de Chardin referred to the fusion of biological life, human culture, and technology as the noosphere. Technological improvement is causing the noosphere to evolve rapidly, driving the enormous increase in human population over the last 10,000 years and the transformation (and devastation) of the biosphere. The rapid proliferation of the internet is changing human culture, including everything from the way we find mates to the way democracy functions, or fails to function. The emergence of the BINC (Bio, Info, Nano, Cogno) technologies promises to further accelerate this change. We are acquiring an ever-increasing ability to engineer devices at a molecular level, to control the genome, and to create new forms of life and intelligence…

First, I believe we’re entering an era of synthesis of modeling. Over the past 20 years, we’ve seen the proliferation of many isolated complex systems models. I think we now need tools for researchers, policy makers and the public to share models. Sharing could happen by stacking different layers of spatial agent-based models in geographic information systems and projecting interactive visualization out onto shared surfaces. Further, we need to make model authoring tools much more accessible to the point where motivated policy makers can author on their own. With the increased ability to author and share models, I believe this will allow us to scale our research to understand and manage the many interacting systems that make up our complex world…

With all the big data available, would it be possible to build a crystal ball allowing us to see everything that is going on in the world in real time? Such projects are in fact under way, built by the military and research centers around the world. These are also political projects, because knowledge is power. Will so much data enable the ruling of a wise king or a benevolent dictator? Could society even be run like a giant machine? Indeed, there are companies that work on such concepts, for example, Google and IBM. They aim at building an operating system for our society that would try to steer our decision-making and behavior with personalized information…

The world consists of inter-connected processes. It is an illusion to think that things exist. As argued by Alfred North Whitehead processes are ontologically fundamental. Since the building blocks do not consist of things with specific intrinsic properties, we can hope that a general science of the processes underlying and controlling the behavior of apparently very different situations, such as the evolution of an ecosystem or the performance of a piece of music by a band, may very well exist…

There really is no science of complexity. Rather we have a fairly well developed set of tools to examine diverse complex and complex adaptive systems. These tools include now familiar ideas of nonlinear dynamical systems, bifurcation theory, and stochastic models, as well as agent-based models such as BOIDS. These tools have been well developed in the past 30 years and we are now underway with the applications of such tools. As B. Arthur noted in analogy, the railways in Britain caused a surge in their stock values, which then fell as the bubble burst, but most of the track was laid afterwards. So, too, complexity burst upon the scene in the late 1980s, largely at the Santa Fe Institute. If that messianic era is now, naturally, past, we are enabled to lay enormous tracks as we proceed…

We live in a world of demographic explosion, bloody ethnic and religious wars, migration galore, increasing inequality and crime, global financial crises, and the danger of pandemics…

For a system to be more than the sum of its parts, it must have strong, long-range correlations, otherwise it could be divided into essentially independent subsystems. In this sense, complex systems are nonlocal. If the constituent parts of such a system are not uniform, the system itself will depend on a high number of details, it will be irreducible, its description will need a large number of parameters, where even tiny details may matter and fundamentally alter the behavior of the system. Such systems cannot be described by a limited set of explicatory variables, they are intrinsically high dimensional…

I want to begin these proceedings by giving some prominence to an elemental tension in the construction of creative institutions. One origin of tension is described by the pragmatist philos opher Charles Peirce: I do not call the solitary studies of a single man a science. It is only when a group of men, more or less in intercommunication, are aiding and stimulating one another by their understanding of a particular group of studies… that I call their life a science…

The last few years have seen the emergence of the sharing economy. As social media blurred the distinction between author and reader, everyone can now offer or receive services thanks to the networking tools provided by new technological companies. Take Uber, and its billion of journeys in 2015 alone, with tens of thousands of vehicles crawling every moment in the globe’s biggest cities. As often, when confronted with a technological change, we observe a polarization of society, and the search for an equilibrium characterized by new norms, rights, and obligations. Understanding the mechanisms behind this re-organization requires an integrated, interdisciplinary approach, covering an intricate web of legal, societal, economical, and computational issues which, we believe, could benefit from a complex systems perspective. As a first step, we are currently studying the dynamics of pricing in Uber. In this new de-regulated world, journey prices fluctuate in time depending on traffic but also on the service’s perceived balance of passenger demand and driver supply…

Compared to the physical and biological sciences, so far complexity has had far less impact on mainstream social science. This is not surprising, but it is alarming because we find ourselves in the midst of a planetary-scale transition from the Holocene to the Anthropocene. We have already breached some planetary boundaries for sustainability, but those tipping points are nearly invisible from the perspective of the linear equilibrium models that continue to hold sway in social science…

Complexity is a highly interdisciplinary science. Although there are drawbacks for researchers to work at the interface of different fields, such as the cost to set up common languages, and the risks associated with not being recognized by any of the well-established scientific communities, some of my recent work indicates that interdisciplinarity can be extremely rewarding. Drawing on large data sets on scientific production during several decades, we have shown that highly interdisciplinary scholars can outperform specialized ones, and that scientists can enhance their performance by seeking collaborators with expertise in various fields. My vision for complexity is based on the added value of its interdisciplinary nature. I list below three research directions that I am personally eager to explore, and that I think will be among the main challenges of complexity in the next 10 years…

In his famous historical account about the origins of molecular biology Gunther Stent introduced a three phase sequence that turns out to be characteristic for many newly emerging paradigms within science. New ideas, according to Stent, follow a sequence of romantic, dogmatic, and academic phases. One can easily see that complex systems science followed this path. The question now is whether we are in an extended academic phase of gradually expanding both theoretical and practical knowledge, or whether we are entering a new transformation of complex systems science that might well bring about a new romantic phase. I would argue that complexity science, indeed, is at the dawn of a new period – let’s call it complexity 3.0. The last academic phase has seen the application of complex systems ideas and methods in a variety of different domains. It has been to a large extent business as usual…

Our societies are being thoroughly transformed by the pervasive role technology is playing in our culture and everyday life. Nowadays the term techno-social systems is adopted to quickly refer to social systems in which the technology entangles, in an original and unpredictable way, cognitive, behavioral, and social aspects of human beings. This revolution does not come without a cost and in our complex world new global challenges always emerge that call for new paradigms and original thinking: climate change, global financial crises, global pandemics, growth of cities, urbanization, and migration patterns. In this framework we progressively face the need to increase the number of people able to imagine original and valuable solutions to sustain large human societies safely and prosperously…

Society currently generates a gargantuan amount of new data each day and a significant amount of these data can be described and modeled in terms of networks and/or flows in them. One ubiquitous character of complex systems is the heterogeneity of their components, of their relationships, and of their pair similarities. To go beyond the detection and modeling of heterogeneity, it is highly informative to filter out features and relationships that cannot be explained by a random null hypothesis taking into account the heterogeneity of the system. Information filtering performed on networks and, more generally, on complex systems allows researchers to detect and characterize structures and phenomena that are present in the system of interest…

In the last 20 years or so, the field of complexity science has entered a new age. The combination of new theoretical insights and the data revolution has prepared the ground for a number of conceptual milestones in many disciplines as diverse as biology, physics, engineering, and economic and social sciences. At the same time, we have been able to identify new challenges whose solutions will confer the science of complex systems an unprecedented applied dimension. Here I would like to focus on one of these challenges: the socio-technical man. With the ever-increasing growth of both the world population and new technologies, it is fundamental for the well-being of humanity and our society to understand how humans interact among them and with the new technological environment…

My vision of complexity sciences targets their potential to extend the range, precision, and depth in making predictions. While this has always been the ambition and yardstick for the physicalmathematical sciences, complexity sciences now allow to include society and social behavior – to some extent. There is agreement that society is a complex adaptive system, CAS, with a few peculiarities. Ignoring, downplaying, or naturalizing them, i.e. to take them as essential and given, carries the risk to end up with abstractions which are cutoff from the dynamics of societal contexts. One of the peculiarities of society as a CAS is that the models with which we try to make sense of the world are invented and constructed by us. It is humans who make observations and provide the assumptions on which models are based. Humans leave traces that are collected and processed to be transformed into data. Humans decide to which purpose they will be put and how they will be repurposed. Humans are object of research and subject. Coping with these peculiarities requires an inbuilt reflexivity. Practioners must perform a double act and do so repeatedly. They must engage in a focused way with their scientific work and equally distance themselves by critically reflecting their often tacit assumptions. A friend of mine, Yehuda Elkana, called this twotier thinking…

My vision for complexity is in the first place visual and has the honor of covering this book. I tried to find an image that represents both the problem and the potential for a solution…

In our societies, we invest a lot of resources to ensure technological progress, and to find out secrets about the deepest corners of our universe. As a result, we are experiencing technological breakthroughs at an unprecedented rate, and we invest billions into research that is as far away from everyday life as the east is from the west. While technology has the potential to improve the quality of life and the satisfaction of curiosity is rewarding, we must not fail to notice that, at the same time, many of our societies are falling apart. Inequality all over the world has become staggering, and we are seriously failing to meet the most basic needs of the majority of people that live on this planet. We have neglected adverse societal changes for far too long, and we have done very little in terms of research to understand, and more importantly, to reverse, or at least to impede, these very worrying trends…

The immense accumulation of data is a new phenomenon which induces many considerations, represents great potential, and sometimes leads to mythical expectations. Here we discuss a specific example of big data applications, the case of economic complexity. This is a new perspective on fundamental economics, adopting a bottom-up approach which starts with a novel use of older data and then develops into its own streamline. The approach confirms some expectations about big data but also disproves others…

Complexity science is for sure a challenging interdisciplinary field of research which will provide new useful perspectives and novel tools for a more efficient participatory society in the next years. Groundbreaking discoveries and innovation are the results of the interplay of many different factors which emerge in a nonlinear and often unpredictable complex way from the fruitful bottom-up mixing of different disciplines. This fact, however, although advocated by many, is unfortunately very rarely fostered and put in practice. Nowadays, funding agencies, often forgetting that research means working at the edge of human knowledge and can be also unsuccessful, rarely tend to risk their money in unconventional proposals regarding complexity. They are more inclined to support scientists, projects, and ideas that have a well settled successful past and operate within well established and more conventional fields of research. This is certainly not a good policy for promoting innovative results. As stigmatized in a recent editorial by Nature (Take more risk, 528, 8/2015), such policies stimulate a conservative and not very efficient behavior, which discourage young scientists and force them to a scientific conformism. The same happens for careers. Today it is not easy for a scientist in complex systems to get a permanent job…

Cybernetics, systems science, synergetics, global systems, complexity - what is new in understanding how the whole works? There are two new opportunities that pose at least two challenges: the availability of massive data and the direct social relevance of the field of research…

Every week, more than 1 million people are currently being added to cities across the globe. This unprecedented trend of urbanization, together with growing concerns over energy supply and climate change, rapidly outpaces existing approaches for the planning and design of cities. A prominent warning example is Beijing’s recent failure to implement a multi-centered urban form that has led to counter-intuitive people flows, immense traffic congestion, and air pollution. Thus, a new quantitative understanding of cities is urgently needed to reduce the risks of such detrimental planning outcomes and to eventually build more sustainable and more livable urban spaces…

Complexity science can help to understand the functioning and the interaction of the components of a city. In 1965, Christopher Alexander gave in his book A city is not a tree a description of the complex nature of urban organization. At this time, neither high-speed computers nor urban big data existed. Today, Luis Bettencourt et al. use complexity science to analyze data for countries, regions, or cities. The results can be used globally in other cities. Objectives of complexity science with regard to future cities are the observation and identification of tendencies and regularities in behavioral patterns, and to find correlations between them and spatial configurations. Complex urban systems cannot be understood in total yet. But research focuses on describing the system by finding some simple, preferably general and emerging patterns and rules that can be used for urban planning. It is important that the influencing factors are not just geo-spatial patterns but also consider variables which are important for the design quality. Complexity science is a way to solve the dilemma of oversimplification of insights from existing cities and their applications to new cities. An example: The effects of streets, public places and city structures on citizens and their behavior depend on how they are perceived. To describe this perception, it is not sufficient to consider only particular characteristics of the urban environment. Different aspects play a role and influence each other. Complexity science could take this fact into consideration and handle the non-linearity of the system…

Definitions of complexity are notoriously difficult if not impossible at all. A good working hypothesis might be: Everything is complex that is not simple. This is precisely the way in which we define nonlinear behavior. Things appear complex for different reasons: i) Complexity may result from lack of insight, ii) complexity may result from lack of methods, and (iii) complexity may be inherent to the system. The best known example for i) is celestial mechanics: The highly complex Pythagorean epicycles become obsolete by the introduction of Newton’s law of universal gravitation. To give an example for ii), pattern formation and deterministic chaos became not really understandable before extensive computer simulations became possible. Cellular metabolism may serve as an example for iii) and is caused by the enormous complexity of biochemical reaction networks with up to one hundred individual reaction fluxes. Nevertheless, only few fluxes are dominant in the sense that using Pareto optimal values for them provides near optimal values for all the others…

My vision is that complexity science will finally become useful. This means we do not only gain groundbreaking insights into the structure and dynamics of complex systems – this goal was already achieved to a large degree in the past three decades. Now, we will turn all these insights, concepts, and methods into something that will improve our socio-technical world – not in a general and abstract manner, but in a detectable and measurable way…

Astronomers of the Maya and babylonian civilizations were brilliant in predicting astronomical events. For instance, from meticulous observations of the sun, Moon, Venus, and Jupiter they were able to predict with astonishing accuracy the 584-day cycle of Venus or the details of the celestial track of Jupiter. Yet they had no clue about our heliocentric solar system, they believed that the earth was flat, and they were completely ignorant of the real movement of stars and planets while being convinced that the sky was supported by four jaguars, each holding up a corner of the sky. if we would be sent back in time and speak to them about the planets orbiting the sun, they would laugh at us and challenge us to come with the accurate predictions that they were able to make. With all our knowledge, but without thousands of years of technological development, we would not be able to come close to any of their predictions. so being laughed at would be a small punishment, more likely we would be ritually slaughtered…

As living beings we as scientists also aim to understand the complications of our surrounding world in terms of models. Models that allow us to forecast future events based on generalization from previous experiences, even when we have not precisely encountered a given situation before. Models may take many forms, being verbal, a painting, a cartoon, or taking the form of a novel with larger ramifications…

Major evolutionary transitions have been described by J. Maynard Smith and E. Szathmary. In this nice synthesis of the history of life, the story ends with the invention of human language. In his book Sapiens, Yuval Harari starts exactly there and summarizes the historical transitions from that point on. If one looks at the complex transitions described by evolutionary biologists and historians one can see that some of those that had the most impact were transitions where the acquisition, analysis, evolution, replication, storage, transfer, and integration of information evolved together with new ways for these information to lead to evolving actions. When we see the technological developments of today that impact simultaneously all these dimensions of information, one is tempted to ask if we are not undergoing a historical and even evolutionary transition. Interestingly, we have the ability to become aware of this very transition, a situation that was not present when the first cells evolved for instance. This awareness may be combined with our understanding of complex systems and the previous transitions to help us make individual and collective choices that may impact our common future…

Understanding the emergent behavior in many complex systems in the physical world and society requires a detailed study of dynamical phenomena occurring and mutually coupled at different scales. The brain processes underlying the social conduct of each, and the emergent social behavior of interacting individuals on a larger scale, represent striking examples of the multiscale complexity. Studies of the human brain, a paradigm of a complex functional system, are enabled by a wealth of brain imaging data that provide clues of how we comprehend space, time, languages, numbers, and differentiate normal from diseased individuals, for example. The social brain, a neural basis for social cognition, represents a dynamically organized part of the brain which is involved in the inference of thoughts, feelings, and intentions going on in the brains of others. Research in this currently unexplored area opens a new perspective on the genesis of the societal organization at different levels and the associated social values…

If physics is the experimental science of matter that interacts through the four basic interactions, the science of complex systems is its natural extension, where the concepts of matter and interactions are generalized. Matter can be anything that is capable of interacting, interactions can be anything that is able to change states of the constituents of a system. Complex systems are made from many constituents (parts) that interact through interaction networks. These parts are characterized by states that change over time. At the same time the interaction networks may change over time. What makes a system complex is that the states of the parts change as a function F of the interaction network (and the states), and, simultaneously, the interaction networks change as another function G of the states of the nodes (and the networks). Physics is about the predictive understanding of the dynamics and changes of states once the interactions and initial and boundary conditions are specified. In complex systems interactions also change over time, and to make things really complicated, these changes are coupled to the dynamics of the state-changes. States co-evolve with the interaction networks. In this sense complex systems often are chicken-egg problems. They are evolutionary, show emergent behavior, can be self-organized critical, show power laws, etc…

Like beauty, complexity is hard to define and rather easy to identify: nonlinear dynamics, strongly interconnected simple elements, some sort of divisoria aquorum between order and disorder. Before focusing on complexity, let us remember that the theoretical pillars of contemporary physics are mechanics (Newtonian, relativistic, quantum), Maxwell electromagnetism, and (Boltzmann-Gibbs, BG) statistical mechanics – obligatory basic disciplines in any advanced course in physics. The firstprinciple statistical-mechanical approach starts from (microscopic) electro-mechanics and theory of probabilities, and, through a variety of possible mesoscopic descriptions, arrives to (macroscopic) thermodynamics. In the middle of this trip, we cross energy and entropy. Energy is related to the possible microscopic configurations of the system, whereas entropy is related to the corresponding probabilities. Therefore, in some sense, entropy represents a concept which, epistemologically speaking, is one step further with regard to energy. The fact that energy is not parameter-independent is very familiar: the kinetic energy of a truck is very different from that of a fly, and the relativistic energy of a fast electron is very different from its classical value, and so on. What about entropy? One hundred and forty years of tradition, and hundreds – we may even say thousands – of impressive theoretical successes of the parameter-free BG entropy have sedimented, in the mind of many scientists, the conviction that it is unique. However, it can be straightforwardly argued that, in general, this is not the case…

The reality of complexity is that causality is very difficult to establish, if at all. Yet we live in a complex world that we seek to manage by establishing causalities. Reality is also that establishing causality is one of the most difficult problems for science, especially for the sciences that deal with the real world. How to untangle or better understand the relationship between causality and reality will be a key in finding ways to sustainably manage our lives, our health care, our laws, and our cities in an ever more complex world…

From schools of fish and flocks of birds, to digital networks and self-organizing biopolymers, our understanding of spontaneously emergent phenomena, self-organization, and critical behavior is in large part due to complex systems science. The complex systems approach is indeed a very powerful conceptual framework to shed light on the link between the microscopic dynamical evolution of the basic elements of the system and the emergence of macroscopic phenomena; often providing evidence for mathematical principles that go beyond the particulars of the individual system, thus hinting to general modeling principles. By killing the myth of the ant queen and shifting the focus on the dynamical interaction across the elements of the systems, complex systems science has ushered our way into the conceptual understanding of many phenomena at the core of major scientific and social challenges such as the emergence of consensus, social opinion dynamics, conflicts and cooperation, contagion phenomena. For many years though, these complex systems approaches to real-world problems were often suffering from being oversimplified and not grounded on actual data…

Over the past quarter of a century, terms like complex adaptive system, the science of complexity, emergent behavior, self-organization, and adaptive dynamics have entered the literature, reflecting the rapid growth in collaborative, trans-disciplinary research on fundamental problems in complex systems ranging across the entire spectrum of science from the origin and dynamics of organisms and ecosystems to financial markets, corporate dynamics, urbanization and the human brain…

The most exciting prospects for complexity science today are in the social sciences. Migration is a good example. According to the UN 720 million people worldwide are currently internal migrants and 120 million are international migrants. How many will there be in 2030, from where and to where do they migrate, why, at what costs and what are the consequences? We require a cross-disciplinary effort involving tools from complexity science, political science, social science, environmental science, psychology, epidemiology, biochemistry, and mathematics to tackle these questions…

Urbanization and innovation are the most defining characteristics of our societal challenge. On one hand, ever-expanding urban built environments are centers of population growth, economic engine, and energy consumption. On the other hand, our technology advances increasingly rapidly, transforming both physical and social infrastructures into better or worse contingency. Therefore, understanding their fundamental dynamics can provide valuable insight into the nature of challenges of sustainability…