By Greg Fisher
Last week the European Commission chose not to invest in FuturICT, which was a massively ambitious project to integrate ICT and complexity science. The aim was, as their website puts it, “understand and manage complex, global, socially interactive systems” and in so doing to “create a paradigm shift”.
Driven mainly by ETH in Zurich and UCL in London, a stellar consortium of universities was created across the entire European Union, and had the active support of MIT in the US. Millions of Euros, provided by the EU, were spent in developing and promoting the concept. The culmination was a bid to be a so-called Flagship EU project, with access to €1 billion – €1 billion! – of funding over 10 years.
Yet the project has failed. A team of highly distinguished but anonymous reviewers, believed to include Nobel Laureates in the social sciences, gave the science of the project a mark below the threshold at which projects can be considered for funding.
What does this mean for complexity science? There is no doubt that it is a serious rebuff. But a key point to note is that it was not rejected at the policy making level. It didn’t get that far. It was turned down on scientific grounds.
There is a view that the project was top-heavy with people from a physics or maths background (all of whom are extremely clever people) who do not always understand that social systems are fundamentally different from physical ones. One MP I spoke to about FuturICT compared it to “The Empire Strikes Back”. I do not mean to kick the academics involved in the project while they are down (I know, like, and get on well with many of them) but I think that in the cold light of day this is a fair criticism.
Extending and reinvigorating the social sciences by complexity science, and in particular by the science of networks, will remain a major theme in the future. This can bring great insights. But the social part of ‘social science’ must never be lost sight of: complex social systems are a very different beast in comparison to natural complex systems.
I feel that Synthesis and organisations like it, are now even more important for the complex systems community, as we try to bridge the gap between the science and the real world policy makers.
FuturICT and Social Sciences: Big Data, Big Thinking
19-07-11 Rosaria Conte, Nigel Gilbert, Giulia Bonelli, and Dirk Helbing
Imagine storing all the computational information produced in the world in just one year: you would have a pile of DVDs able to reach the Moon and back. How about all the data collected since the beginning of the computer era? The quantity is so huge that traditional units of measurement cannot cope. For this reason a few years ago computer scientists started talking about “Big Data”, referring to the gathering, formatting, analysing and manipulating of a massive amount of digital information.
For social scientists, the challenge is now to make sense of all these data in a way that illuminates our social world. This will involve a collaboration between social scientists, who have the concepts, theories and analytical expertise that are needed, and scientists and engineers, especially computer engineers, physicists, and complexity scientists, who are used to handling vast amounts of data.
The European Commission, as part of its Framework 7 Programme, has launched a competition for proposals for “Flagship” research initiatives that will make major advances. Six themes have been accepted for further development; of these only two will be selected and then funded with €10 billion over ten years. Among the themes being considered are research programmes in nanotechnology, robots, personalised medicine, and one on understanding complex social systems, called FuturICT.
Unleashing the power of information for a sustainable future
FuturICT aims at understanding and managing complex social systems, with a focus on sustainability and resilience. Its starting point is Big Data: models of techno-socio-economic systems will be developed, grounded in data from existing and new information technology systems. Computational Social Science will play a crucial role: revealing the processes underlying the emergence and maintenance of societies constitutes a major challenge of the project. FuturICT will help the social sciences take advantage of the computational instruments and the data-driven knowledge required for building and testing social science knowledge.
But is Big Data enough? Social scientists working within FuturICT believe it is not. Just because we have billions of bits of information does not mean we know the intention behind them or their consequences. For this reason, the Social Sciences have a vital role in the project in order that it not only has good data but also good theories.
The greatest difficulty for a project like FuturICT lies in asking good questions. For example, why do financial markets crash again and again? How can we construct resilient institutions? What determines human happiness and well-being, and how are they influenced by personal wealth? How can society change behaviours that destroy our environment and other important public goods? These are just some of the crucial social problems FuturICT will work on. After an analysis of these issues and their consequences, data mining and large-scale computer simulations will step in to provide empirical tests. However, “why” questions cannot be answered through data analysis alone. Therefore, fundamental questions will be discussed during so-called “Hilbert Workshops”—think tanks dedicated to triggering new approaches and ideas. These will naturally lead to the construction of novel theories, developed with the help of an innovative ICT, both responsive and responsible.
So, big questions, big data and big theories: FuturICT goes beyond pure information, and takes the path of big thinking.
http://www.futurICT.eu
I think it is somewhat unfair to say that “It was turned down on scientific grounds” and then imply what was wrong was a lack of understanding of social science . Rather oddly, most of the criticism in the evaluation was of the ICT capabilities of the proposal. For instance, the first substantive criticism mentioned in the evaluation was “The theoretical part of the plan is clear, coherent and consistent. The approach is generally sound. However, a weakness of the proposal is that it does not properly anticipate the ICT resources required to achieve the level of large scale data processing that will be eventually required for the world model. There is concern regarding the lack of specification on the ICT requirements of the exploratories – for example it is not clear how many people are participating. Therefore the requirements cannot be determined. Furthermore, the ICT architecture to be developed according to a social model is not sufficiently specified.” The proposal may of course have had social science failings too, but that was not the criticism made by the evaluators.
Thanks Nigel.
When I wrote that it was turned down on scientific grounds, I was referring to paragraph 3 of part 1 of the evaluation document: “… the proposal does not offer sufficient insight into how the collected data and models are to be validated”.
You’ll know better than I that validation is a crucial part of science, as Karl Popper wrote about at great length. Hence I think it is reasonable to state that the FuturICT failed (in part) on the basis of science.
My point about “people from a physics or maths background… who do not always understand that social systems are fundamentally different from physical ones” is pure opinion by me based on what I have heard from a number of (certainly not all) people on the technical side re complex social systems (I don’t include your good self!!). I have heard the same opinion from several other people. I didn’t suggest this view was in the evaluation.
I should be clear about this: there is a role for technical analysis and high-end mathematics in this work, otherwise we’ll be left waving our hands. What I am referring to here is an issue of emphasis.
Having read the evaluation, I hope to find time to criticise the critique because some of the points made sat awkwardly with the notion of planning in large complex projects.
Best wishes,
Greg
I completely agree with your analysis Greg. Complexity approaches should enrich and supplement social theory and policy not attempt to subsume or recast them as a physics problem. For me the key contribution of complexity approaches is to open up new ways of seeing things in a more holistic way and to communicate this in models, simple models, that can be comprehend by people from all relevant groups and backgrounds.
I don’t believe that using complexity techniques leads to any particular agreement on the nature of social reality other than avoiding reductionistic thinking. Moreover, I don’t see that the complexity approach is in any way tied to a technocratic command-and-control view of social policy.
The danger has been that a form of “scientism” has crept in to social complexity circles because the backgrounds of many of those working in area are from the physical, mathematical or computer sciences. Policy has been seen as a technical / scientific exercise in which “experts” will solve problems given to them by policy makers using clever computer models.
I personally think this is not only wrong but dangerous. It leads to a kind of technocratic rationality that is not able to challenge fundamental assumptions and / or reformulate problems in new ways – and this is precisely what is needed right now.
Social reality is a creative and on-going collective construction. It is visible from many perspectives and can be changed from within. There are no “laws of society” analogous to physical laws. The complexity approach can enrich our views by offering new perspectives and new possibilities. We can use computer models to help us to think through computational thought experiments – not to predict the future but to think in new ways and identify new possible futures.
One future I see is not one controlled by the dead hand of a technocratic elite but ever evolving decentralised networks of free flowing information leading to collaboration and cooperation from the bottom-up. Well, you’ve gotta have a dream right?
I enjoyed both the blog, Greg and David’s comments.
“One future I see is not one controlled by the dead hand of a technocratic elite but ever evolving decentralised networks of free flowing information leading to collaboration and cooperation from the bottom-up.”
So not a project designed to attract EU funding, then.
As someone with a Bsc.Hon from UCL but in soppy old psychology, I’ve always had the nagging feeling that if I can’t speak equations I’m not a proper scientist, I wrote a book, but that didn’t help – now i’m like a science journalist. Pah.
This result makes me smile: I’m smiling at most of the parties concerned. Sounds like wrong approach right problem turned down rightly for wrong reason. That may well be what passes for a rational decision in Brussels.
Nice diplomacy though, Greg.
Arthur said:
“Sounds like wrong approach right problem turned down rightly for wrong reason. That may well be what passes for a rational decision in Brussels.”
Brilliant! I believe that the way science is funded and winners are selected is a big issue which we discuss very rarely. There is a kind of taboo on critical evaluation of these procedures within academic circles (at least in public) because quite frankly people are afraid of rocking the boat and losing their funding.
I personally believe that something akin to crowdfunding – with any citizen having a say – could be one way forward. Why not give every tax paying citizen one vote to decide what to fund? Crowdsource the reviews!