Chiara Marletto is a research fellow at Wolfson College, University of Oxford. Her research is in theoretical physics – especially quantum computation, thermodynamics and information theory. Her broader interests include theoretical biology, epistemology and Italian literature. The Science of Can and Can’t: A Physicist’s Journey Through the Land of Counterfactuals attempts to forward a new foundational basis for physics. It is her first non-academic book.
You argue for a radically different approach to physics, which you call the science of can and can’t. What does that mean?
It’s a new mode of explanation. Since Newton, traditional physics has been using laws of motion, describing how objects move in space and time – what happens to an apple if you set it in motion in this or that way. With one exception: thermodynamics. The laws of thermodynamics prescribe the impossibility of perpetual motion; by doing so, they put powerful constraints on all laws of motion – those known and those yet to be known. Constructor theory follows the same logic, but it extends to a much broader context. We express all fundamental laws as constraints about what transformations are possible and impossible. This apparently simple switch is very powerful. For example, it can capture entities that traditional laws of motion cannot handle exactly: information, the physics of life, and even the mind.
For much of the past century a debate has raged in physics about how to reconcile quantum theory and general relativity. Is your book a step towards an answer to that question?
The science of can and can’t is at a deeper level than general relativity and quantum theory. It was proposed by quantum computing pioneer David Deutsch to expand on the quantum theory of computation. Like the latter, it consists of deeper physical principles – guidelines for consistently putting together different laws of motion, such as quantum theory and general relativity, while still preserving their respective core features. So it provides key new tools to help with that question. And that’s good. It’s like with Covid: you want to try all possible ways to solve the problem.
Many scientists have explored complex theories, like string theory, as a means of bridging these two apparently irreconcilable worlds. Why don’t you mention that kind of theoretical physics?
My research isn’t into finding a candidate to merge general relativity and quantum theory, so those proposals didn’t fit the focus of the book. I think one difficulty with them is that they find it hard to provide testable predictions. The science of can and can’t is helpful here, because it provides new paths to testable predictions, even in domains relevant to quantum gravity.
You talk about the “counterfactual” in the book. What do you mean by that term?
The way I think about counterfactuals is specific to physics. Counterfactual statements refer to what is possible or what is impossible, as opposed to what happens. Take Heisenberg’s uncertainty principle: it’s impossible to build a perfect measurer of both position and velocity for an electron. It’s not about the fact that a perfect measurer will not happen given a particular initial condition; Heisenberg says that it can’t happen at all, no matter what the initial condition. That’s a much stronger requirement.
Would you say your book was as much about the philosophy of science as it is an argument for a particular kind of physics?
It has lots of philosophy in it. I would say the best physics arguments are very philosophical. But the book is primarily about physics.
You argue against reductionism, the idea that everything in the universe can be reduced to the dynamics of elementary particles. Is there any danger of the supernatural creeping into an anti-reductionist conception of things?
I think there’s this misconception that the only way to remove the supernatural from our explanations is to reduce everything to microscopic dynamical laws and initial conditions. That is simply one possible level of explanation. But there are other things that are also explainable in scientific terms, without appealing to the supernatural, but they can’t be reduced to that level of explanation. An example is the laws of computation: they aren’t microscopic laws of motion, but they are compatible with them. They’re not saying that computers are magic: they are physical laws that capture some phenomena in nature, at a different explanatory level. If you stick solely to microscopic laws, you will miss those regularities in nature that allow for classical and quantum computers.
What is the relationship between quantum computing and quantum theory? Which is leading the other?
Quantum theory came first, historically. But in the 80s, some scientists realised that the perplexing aspects of quantum theory actually made a lot more sense in relation to computation theory. It turned out that by studying the properties of the universal quantum computer – a theoretical development of the Turing machine – we could actually understand quantum theory much better. So I regard the theory of quantum computation as more fundamental because it captures the (counterfactual) foundations of quantum theory.
What do you imagine might be the practical applications of the science of can and can’t?
The most spectacular application is the universal constructor, a machine that can be programmed to perform not just all possible computations, but all physical transformations that are allowed by the laws of physics. It’s an all-powerful 3D printer. There could be an era, far in the future, where the universal constructor is part of our lives as computers are now; this could revolutionise our civilisation.
You’re talking about extraordinary computational power. Are you anxious about where such power might lead as we enter the age of artificial intelligence?
Yes, I think as a scientist you do have these worries. But I place great value in the knowledge that humanity can create, and what I’m hoping is that as we make progress in science, our society can also make progress to look out for potential problems that might come out of new science applications, and solve them.
How far away are we from quantum computers becoming reliable working computers?
Closer than ever, but we don’t know how far. The initial goal of building the universal quantum Turing machine has been subdivided into smaller goals of building special purpose quantum computers that can address specific tasks, one at a time – cryptography is an example. These quantum technologies are already here and can be used in all sorts of fields, from engineering to biology and medicine.
Physics is traditionally a very male-dominated field. Was that a discouraging image when you were deciding to enter the field?
When choosing physics I didn’t feel that there was a problem, because my parents empowered me. They always treated me as an individual, so the way I interact with the world is not primarily through the fact that I happen to be a woman. I’m a scientist and I’m interested in physics, that’s all. Sometimes I’ve encountered residual incredulity about the fact that there are women in physics. But if we persist at it, a culture shift will occur. Girls will then realise they can do anything they want (provided it’s allowed by the laws of physics).