Why interpretations are not useful (to understand quantum mechanics)
TL;DR: Understanding quantum mechanics means understanding why the model applies to objects at different scale
When it comes up that I work on the foundations of physics, people assume that I am interested in interpretations of quantum mechanics, and they are surprised when I say I am not. To be fair, I was very curious about them initially, but in time I came to the conclusion that they didn’t really help me better understand the theory, and they are actually a distraction.
By interpretations I strictly mean those accounts of quantum mechanics that do not change the actual predictions. I consider the others, those that give different predictions, different theories, as they are experimentally distinguishable from quantum mechanics. Interpretations, on the other hand, try to give a “deeper” account of what happens in the real world, especially during a measurement, often by positing elements that are not experimentally accessible. For example, they may claim that there exist “hidden” variables, experimentally inaccessible quantities, that would give a complete description; or claim that, during a measurement, the universe itself branches off into different, separate, universes where each possible outcome is realized; or claim that “consciousness” causes measurements to happen; or that it’s all about agents tracking subjective information and so on.
In my university years I found all this fascinating, and why wouldn’t I? But once the novelty wore off, I realized I was no better off than when I started, which is probably the experience of most physicists and engineers. The topic is just a never-ending debate, given that these questions cannot, by definition, be settled experimentally. But worst of all, I was no closer to answer even the most basic questions: what is a quantum system?
Being an engineer at heart, I see physical theories as models that you use when certain conditions apply. For example, one can use the theory of linear circuits for electric, hydraulic, thermal or mechanical systems, as long the current is not changing over time, the water flow is laminar and so on. There is no intrinsic ontological property associated with a linear circuit: as long as the response of the system, whatever it may be, can be assumed to be linear, the model applies. Even when studying physics, you are essentially taught a bag of different tools, one to be used for cannonballs, one for heat engines and so on.
In this light, quantum theory is just another model that can be used to describe some objects in some conditions. We don’t use it to describe cannonballs or heat engines, but we can use it for electrons, protons, atoms and even buckyballs, which are molecules of carbon shaped like footballs (the proper ones… not the American ones). In the appropriate conditions, all these systems will behave like a single quantum system: they will diffract, interfere with themselves and so on. Understanding quantum theory means, at the very least, being able to tell what makes those systems, in principle so different, amenable to the same description.
An interpretation that focuses on underlying mechanisms, then, has the problem that the mechanisms cannot possibly be the same for all these disparate systems. What happens during an energy measurement of an electron will be very different from what happens during a position measurement of a whole atom. Conversely, an interpretation that focuses on the agent who performs the experiment needs to justify why in those cases the quantum rules apply, rather than the classical ones. When we brought up these issues to those working on interpretations, the implicit or explicit conclusion was always that they were only interested in fundamental objects. But this begs the question: since we use these tools for objects we know are not fundamental (i.e. a buckyball is made of sixty carbon atoms, each made of six electrons and twelve nucleons, six protons and six neutrons, each made of quarks and gluons), why can we do that? Why can’t we do that for a cannonball?
If you are looking to truly understand quantum mechanics itself, when it applies and what properties is it describing, quantum interpretations will not help. We found it more helpful to step back and reflect on what exactly a physical system is to begin with. This seems like a prerequisite for deciding which description is appropriate. It is a more fundamental question.