Can re-purposed science help us understand more than the physical world? Rachel Brazil talks to the scientists trying to play swap
Chemistry is underpinned by 200 years of chemical theory. The subject’s theoretical basis explains and mathematically models chemical processes and allows chemists to make predictions as to how molecular systems will behave. But in recent years, hunting for inspiration and predictive power, the social and political sciences have started to borrow concepts from the physical sciences to model very different phenomena. Sharing theoretical frameworks is an exciting new move for some from both sides of the science/social sciences divide, while others are much more sceptical. Can models developed to describe the physical world really explain and predict how people and economies work? And what about the other way round?
Entropy and ecological economics
Developed in the 1980s, ecological economics has gained traction in a small and enthusiastic community of economists. Their field is based on the idea that just as the physical world does, the economic systems must obey the law of entropy. This concept came from Romanian American economist Nicholas Georgescu-Roegen who published The Entropy Law and the Economic Process in 1971.1 It is still relatively marginal to mainstream economics. ‘It has in its own right actually created a kind of momentum, a critical mass,’ says Tim Jackson, director of the Centre for the Understanding of Sustainable Prosperity at the University of Surrey in the UK.
The thermodynamic concept of entropy was developed by German physicist Rudolph Clausius in 1865, from work on the efficiency of heat engines. He concluded that in energy conversions, the ability to do work or the ‘quality’ of the energy is always downgraded – ie entropy always increases, now known as the second law of thermodynamics. For chemists, entropy is often equated to ‘disorder’ and attributed to the number of different molecular arrangements available in a given system. An increase in total entropy is one of the factors that indicates the spontaneous direction of a reaction.
Eventually, the store of energy is going to run out
Jackson explains that Georgescu-Roegen’s use of entropy also stems from the work of Russian-born Belgian chemist, Ilya Prigogine, who won the 1977 chemistry Nobel Prize for his work on far from equilibrium thermodynamics. Head of what became known as ‘the Brussels school’, Prigogine explained the emergence of complex order (such as life) through the availability of high quality energy, which must be accompanied by disorder in the surrounding system.
Ecological economics uses these ideas to explain what is and isn’t possible in economics. Georgescu-Roegen believed that the economic process is subject to the second law. Every industrial process irreversibly transforms the earth’s resources (energy and matter) into products but also always into waste, such as carbon dioxide. ‘There will be thermodynamic consequences to that, and eventually that store of energy is going to run out,’ Jackson explains.
The conclusion is that economic growth cannot be infinite, with the term ‘entropy pessimism’ being coined. ‘It gave rise to a sense that the internal logic of a growth-based economics was inappropriate on a finite planet that was governed by the laws of thermodynamics,’ says Jackson. Some in the field have taken the analogy even further and perform ‘exergy’– cost analyses on industrial processes, in order to evaluate their impact on the current natural environment. Exergy is defined as the available energy for conversion to work from an energy source.
Many economists reject this use of thermodynamics, calling it pseudoscience. ‘In traditional economics there isn’t a great understanding of entropy,’ says Jackson. ‘There is a lot of confidence in the idea that it is possible to decouple physical and material things from economic things.’ They argue that the Earth does not represent the kind of isolated systems in which the concept of entropy was originally defined or that it’s simply not possible to use thermodynamics this way. Jackson rebuts this. Despite the Sun providing renewable energy, to be useful it needs to be captured and concentrated, a process that itself needs resources and means any economic system must exist within the bounds of this solar inheritance.
The need to conserve the Earth’s resources seems like a common sense argument, without need for thermodynamics. But the weight of a scientific theory is something many disciplines crave. ‘Knowledge made by physics, biology and chemistry is considered by many to be more authoritative, more solid, more convincing and powerful,’ explains Stephen Kellert, a philosopher of science at Hamline University in Minnesota, US, who has written extensively on discipline-hopping theories.2 ‘[By] borrowing knowledge, ideas, concepts, techniques and terminology, you gain something of the aura, status and prestige of those fields,’ he suggests.
Modelling opinions and political change
Another area to make a claim on scientific theory is belief dynamics – understanding how points of view change and spread in political or social debates. The growing field of socio-physics is using statistical mechanics to model these processes. Father of the field, Serge Galam, is now based at the Paris Institute of Political Studies in France but started working on the area as a physics student. He found social science oddly uninterested in creating theories with predictive power. ‘I got more and more convinced that we needed to develop a new methodology to tackle social systems the way physicists do,’ says Galam.
The theory Galam and others have borrowed is the Ising model, developed in the 1920s to model ferromagnetism. The Ising model is named after the physicist Ernst Ising and it models interactions between atoms in a lattice, each with their own spin, in one of two states, and associated magnetic dipole moments. The spins of neighbouring atoms interact, with each spin energetically pushed to align with its nearest neighbours and the whole system driven towards an ordered magnetic state. But above a critical temperature, thermal effects cause spin fluctuations and destroy this order.
Physicist Daniel Stein from New York University in the US explains why modelling how political opinions change has similarities to feromagnetism. ‘What is of interest is not just the properties of [people] themselves but also the interactions between [people] and this is where the statistical mechanics comes in, because it’s all about the interactions [between] individual particles and spins.’ So as each atom influences neighbouring atoms, humans influence others in their networks, be they friends, family, experts or media figures.
In August 2016, he suggested Trump could win the November general election
Social psychologist Mirta Galesic from the Santa Fe Institute in New Mexico, US, is working with Stein to model how interactions within social networks cause people to change their opinions – such as voting for a left- or right-wing candidates, or beliefs on issues like vaccination and GM foods.3 They used a form of the Ising model, with people’s opinions represented as up or down ‘spins’ and changing in line with their coupling to surrounding ‘spins’ – which represents the opinions of the network of people with which they come into contact. Their model can also include elements such as a person’s initial political leaning. ‘In the random-field Ising model every particle can have a different internal magnetic field which in our case is an analogy for all the beliefs and values that you have from before,’ says Galesic.
Within mainstream social psychology this sort of work goes largely unnoticed, but Galesic says the typical reaction would be to say ‘This doesn’t make any sense, people are so much more complicated than physics and particles.’ One difference is the type of networks formed. ‘In statistical mechanics models, one uses a network like a Euclidean lattice,’ says Stein which doesn’t represent how people’s networks are organised. There are some models that can better simulate social interactions such as the small world network which creates clusters and results in strangers being closely linked (think of the ‘six degrees of separation’ concept).
Galam has used his model to explain and predict unexpected political outcomes such as Brexit and Trump. Crucially his model is able to predict tipping points – in the Ising model, this will cause a phase transition from a ferromagnetic to a disordered state; in opinion dynamics, from one election result to another. These tipping points explain why a vocal minority can turn a majority towards its position. In August 2016, he suggested Trump could win the November general election – a view not then supported by analysts or polls.4
Galam takes into account the proportion of entrenched ‘stubborn’ voters who will not budge their opinion but who are able to influence others. He also throws in what he calls hidden prejudices and explains that when people are uncertain over a choice they will be influenced in line with their unconscious biases (he says this explains the shift towards Trump and Brexit). The upshot is that his model can predict the tipping point at which one view will be able to eventually dominate. In some scenarios, a position with an initial support of only 17% can shift opinion over a public debate and end up winning a majority.5
Using models from the physical science seems to bring a new understanding of some fundamental features of social and political systems. ‘My work is disturbing in the sense that it shakes [our] linear theories about democratic debate and referendums,’ says Galam. Winning a public debate or election can apparently be achieved not by convincing a majority of people, but by increasing the proportion of entrenched voters on one side who can influence others.
Chemical game theory
Darrell Velegol, a chemical engineer at Penn State University started thinking about re-purposing scientific theories as a student. In a political science assignment on why Americans don’t vote, Velegol remembers framing his argument in terms of Ohms law. ‘On that section [the professor] wrote “Nice”, and I have been thinking about it ever since.’ As an academic Velegol turned his attention to what he calls contested decision making – the kind of problems solved using game theory. He has devised a whole new framework which he calls ‘chemical game theory’ (CGT). ‘I came to see analogies with chemistry and chemical engineering that I hypothesised might explain things a little better,’ he says.
If the model never explains the data, maybe the model is irrational
A traditional problem in game theory is the prisoners’ dilemma. Two members of a criminal gang are arrested and imprisoned, each placed in solitary confinement. If both prisoners remain silent they can only be prosecuted for a lesser charge, but prosecutors offer each prisoner a plea bargain – if they betray the other prisoner they will receive a more lenient sentence. The prisoner must decide whether to betray their partner, or keep quiet in the hope that the other prisoner does the same, but running the risk of being themselves betrayed.
Traditional game theory states that the best solution is always to betray your partner-in-crime. But in reality this doesn’t happen. Theoretical experiments on groups of students show that they select this solution only 50% of the time. ‘Economists usually just say that people behave irrationally. Maybe its because I’m in chemical engineering, I think OK, if the model essentially never explains the data, maybe the model is irrational,’ says Velegol.
Using chemical game theory he has come up with a model for real decision making behaviour.6 His model frames the decision making process as a series of competing chemical reactions with an energy of reaction calculated for each potential decision or ‘reaction product’. ‘That’s where our friend Gibbs comes in. He told us how we could solve these chemical reactions using thermodynamics,’ says Velegol. ‘We can start thinking about it just like you would in a chemical reactor.’ He uses the energy of each equilibrium ‘reaction’ thermodynamically, finding the Gibbs free energy of all possible outcomes and then working out the scenario that gives the lowest overall Gibbs free energy to find the fractions of each potential reaction or decision. This gives you the players’ strategies in terms of the likely proportions of multiple ‘products’, rather than focusing on only one solution.
The model can also account for real world factors such as players’ preconception and attitudes – for example, your solution to the prisoners’ dilemma may differ if the other prisoner is a member of your family. Velegol uses initial concentrations to stand in for pre-existing biases that might make one decision more likely than another.
Chemical game theory is still a hypothesis, and Velegol says it needs much more testing, but so far it has been able to replicate prisoners’ dilemma results. Using chemistry as an analogy seems to provide a model for the complex and varying responses of humans. But as Velegol says, to compare predictions with some real experiments to test the hypothesis; ‘we have a long way to go in this area’.
More than a metaphor?
All these examples use scientific theories to explain different phenomena, but is this any more than a colourful metaphor? ‘There is a fine line between using a model as a framework to organise knowledge and interpreting this too literally,’ says Galesic. Galam agrees: ‘The point is not to map physical theory so directly to social problems,’ he says. But the analogy works because both systems represent collective phenomena which emerge from local interactions – between particles in physics and between people and their networks in opinion dynamics.
Kellert says we shouldn’t underestimate the power of a metaphor. ‘The history of science shows us that metaphors are actually incredibly powerful … metaphors can also fundamentally reorient the way you think about your field.’ Though there has been push back against the misuse of scientific concepts, which is epitomised by the well-known hoax from physicist Alan Sokal from New York University in 1996. He submitted a spoof article on quantum gravity to a cultural studies journal, which duly published it. ‘It’s very distasteful to see people with a very rudimentary understanding run wild with metaphors in a way that’s not especially enlightening, but it doesn’t always have to be that way,’ says Kellert.
I am somewhat suspicious of claims about universal insight
In ecological economics, Jackson says the idea of entropy is supposed to be taken not just as a metaphor but as a physical reality. The waste we can see being created from economic activity is a consequence of the entropy law. ‘It’s something that’s more empirically founded, I would argue, than necessarily theoretically founded.’
Velegol says that his work started off as purely metaphorical, but now he has started to see it at something more. ‘You start to see common languages used to describe many different phenomenon and I think chemistry and chemical engineering is a way of describing information more generally, even beyond atoms and molecules … we started this purely as a metaphor but we think it can be extended and who knows where it will lead.’
Whether theoretical models from physical science can provide universal insights into all fields is an open question. ‘I am somewhat suspicious of claims about universal insight or grand multidisciplinary syntheses – sometimes they work and sometimes they don’t,’ says Kellert. ‘We can’t rule it out,’ he adds.
Many people are sceptical or outright dismissive
Those working on these borrowed theories are convinced they can provide real value to society. Velegol hopes to use chemical game theory to explore the classic ‘tragedy of the commons’ problem, where individuals acting according to their own self-interest behave contrary to the common good. The idea was coined in 1833 to describe the degradation of public grazing land. ‘Of course the big one now is climate change,’ says Velegol. ‘We want to explore that problem and see if there are creative solutions.’ He is also looking at innovation and how companies decide on future product directions. If his work can make decision-making 2% better, he says it would still make a big difference.
Meanwhile Galam hopes sociophysics will improve our understanding of politics. ‘I think one day we will be able to predict on solid ground some co-operative phenomena, like an election or a social movement, for a few weeks or a few months.’ The approach may even provide an early warning of increases in extreme or fringe views that are in danger of spreading to the wider population.
‘The physical sciences are a treasure trove of really interesting and powerful concepts, techniques and methods,’ says Kellert. Other disciplines have been taking notice and by borrowing these theories are able to generate genuine insights. ‘Many people are sceptical or outright dismissive,’ says Velegol, but ‘on the other hand you will find many people who are deeply intrigued with the idea. You have this cohort of chemists and economists, physicists and ecologists thinking this is important and pro-actively working on it.’
Rachel Brazil is a science writer based in London, UK
1 N Georgescu-Roegen, The Entropy Law and the Economic Process, Harvard University Press, 1971 (DOI: 10.4159/harvard.9780674281653)
2 S H Kellert, Borrowed Knowledge: Chaos Theory and the Challenge of Learning across Disciplines, The University of Chicago Press, 2008
3 M Galesic and D L Stein,Physica A, 2019, 519, 275 (DOI: 10.1016/j.physa.2018.12.011)
4 S Galam,Int. J. Mod. Phys. B, 2016, 31, 1742015 (DOI: 10.1142/S0217979217420152)
5 S Galam,Physica A, 2010, 389, 3619 (DOI: 10.1016/j.physa.2010.04.039)
6 D Velegol et al,Ind. Eng. Chem. Res., 2018, 57, 13593 (DOI: 10.1021/acs.iecr.8b03835)