In a trio of blog posts from 2010 (see here, here, and here), Sean Carroll defends the striking claim that, as far as concerns the basic physical principles that underlie the phenomena of everyday life, physics has been completed.
[T]here’s no question that the human goal of figuring out the basic rules by which the easily observable world works was one that was achieved once and for all in the twentieth century.
That’s right: “once and for all.” If asked for the basic, underlying story about why a table is solid or why the sun shines or what happens when a person flexes a muscle, modern science gives its answers in terms of “the particles of the Standard Model, interacting through electromagnetism, gravity, and the nuclear forces, according to the principles of quantum mechanics and general relativity.” One hundred years ago, explanations by this story (i.e., body of theory) could not be given, because this story did not exist. “But—here’s the important part—one thousand years from now, you will hear precisely that same story.”
I think Carroll is right, and I think the philosophy of structural realism can help to illuminate why. The purpose of what follows is to explain these points.
To begin, it is necessary to be clear about what Carroll is and is not saying. He is certainly not saying that physical theory is complete. For example, we have no quantum theory of gravity. We do not have a unified theory of the forces of nature, which we would like to have. Again, quantum field theory is perturbative and therefore inexact (albeit extremely accurate), which is less than ideal. Furthermore, there are many phenomena that current physical theory cannot adequately explain (or which even contradict it), such as those having to do with dark matter, dark energy, the Higgs potential that induces spontaneous symmetry breaking, matter/antimatter asymmetry, and the fact that quantum field theory predicts an energy for the vacuum of empty space that is 122 orders of magnitude higher than its observed value (amusingly referred to by physicists as the “vacuum catastrophe”). Therefore, theory in fundamental physics is very far from completed, and everybody knows this, including Sean Carroll. Nor is he saying that existing physical law can explain every exotic fact we might wish to explain, such as gravitational lensing or the accelerated expansion of the universe or why there are three and only three generations of fermions. These are facts it will take new theory to explain.
Instead, Carroll emphasizes that he is talking about the phenomena of “everyday life,” by which I think he means to include not only the solidity of tables and the flexing of muscles, but also advanced technology and anything that can make a practical difference in our lives. For instance, he mentions, as a possible exception to his claim, high-temperature superconductivity, for which there is currently no explanation. But his reply isn’t that it’s not part of everyday life, but rather that its eventual explanation will not require new theory: “Nobody thinks we’re going to have to invent new elementary particles or forces to understand high-Tc superconductivity.” This illustrates two important points about his claim. First, he is not saying that all the phenomena of everyday life are in fact understood. We do not fully understand the weather, cancer, photosynthesis, mental phenomena, or economics. But understanding them is not going to require any new fundamental physics. This is the second point: The physics we have now will be adequate to explain all the phenomena of everyday life. (To the extent that it is the role of physics to explain it. Carroll can sound naively reductionistic sometimes, but let us set that aside. I am also setting aside his claim that physics will explain consciousness. On this point, he is not naïve, just wrong in my view, as my post linked earlier would indicate. There are important questions raised here, but they distract from the main argument. Neither of these issues changes the basic truth he is pointing to, which is that “the physics of everyday life is completely understood.”) And the criterion of everyday life is basically: whatever on Earth we can observe without having to construct a billion-dollar particle collider to observe it.
Thus, Carroll’s claim is limited to what he calls the “everyday regime,” which he does not crisply define but which seems to come down basically to events within certain limits of energy, duration, size, acceleration, etc. And his claim is that we know the laws that govern physical events within that regime “once and for all.”
However, he gives little by way of argument for it. His justification seems to be that physicists generally think it is true, although they rarely state it explicitly, and that he can’t think of any counterexamples. But I think we can do a little better than that by appeal to the principles of structural realism.
Structural realism, as I conceive it (and I promise that’s the last time I will link this piece), is an epistemological doctrine. (There are other interpretations, many of them unfortunately quite absurd in my opinion. See here for a review.) It says that all we will ever learn about the world outside our own minds is its structure and dynamics (or—I think equivalently—its topology and causal interactions). The reason is that to learn about anything outside of your mind requires that you causally interact with it, so that what can make no causal difference cannot be known. Thus, any non-relational, purely intrinsic properties of things outside the mind are forever inaccessible to it. Of course, unless one believes the world could consist of relations all the way down, there must be some purely intrinsic properties of things outside the mind; only we will never know what they are. The function of consciousness is to supply from the inside a substitute for what we cannot learn from the outside. Hence, colors, tastes, felt hot and cold, sounds, etc. are the terms in which we perceive the world, but there is no reason to suppose that they really exist in the world. This is the form of structural realism advanced by writers such as Russell, Maxwell, and Blackburn. It is supported by various converging lines of evidence, among which I would particularly mention David Lewis’s approach to defining theoretical terms.
Now, the approach that I just stated is sometimes called the “upward path” to structural realism. It starts from epistemological principles and works from them to conclusions about what we are able to know. It is contrasted with the “downward path,” associated with the history and philosophy of science, which finds that a staunch realism concerning the interpretation of scientific theories must be relaxed to a weaker, “structural” form to achieve adequacy. The locus classicus of this form of structural realism is a well-known paper by John Worrall. I think Carroll’s claim is best understood by means of the downward path to structural realism, so let me say a little bit about it.
Worrall presents structural realism as “the best of both worlds,” the two worlds being realism and anti-realism. Realism, in Worrall’s view, has the virtue of making sense of the success of scientific theories: they are successful because they are true! Heliocentric astronomy succeeded because the earth really does go around the sun, the germ theory of disease succeeded because there are germs, the atomic theory succeeded because there are atoms, etc. The problem with this is that there are, not uncommonly, scientific revolutions in which theories—including quite powerful and empirically successful ones—get replaced by new theories with very different ontologies. An example would be the replacement of Newtonian mechanics by general relativity. It is sometimes said that general relativity subsumes Newtonian mechanics as “a special case,” but this is a misconception. It is true that the equations of Newtonian mechanics can be derived from those of relativity under certain (never realized) limiting conditions. But this neglects the fact that the two theories make radically different claims about the constitution of the world. Newtonian mechanics pictures space as an absolute, infinite, Euclidean container in which objects that have mass instantaneously attract each other at indefinitely large distances, and in which time, size, and mass are constant properties of objects independent of their velocity. Famously, not one part of this long statement is true in relativity theory. Therefore, although the two theories share some of their mathematics, their descriptions of the world are utterly different. Thus, the scientific revolution that ushered in the theory of relativity was not merely an adjustment or updating of Newtonian mechanics, but the wholesale rejection of the Newtonian description of the world and its replacement with a very different description. And if this can happen to so successful a theory as Newtonian mechanics, what theory is safe? And if no theory is safe, where does this leave the argument that the empirical success of a scientific theory gives us the right to infer that its description of the world is true?
Here, Worrall’s insight (for which he credits Poincaré) is that we can distinguish the relational structure of a scientific theory, as embodied in its mathematical laws, for example, from its ontological claims and then note that although the ontology of an old theory may be wholly discarded by the new theory during a scientific revolution, the relational structure is carried over in an approximate way. Most commonly, the survival of the old theory’s relational structure shows in its appearance in “limiting cases” of the new theory. Thus, as already noted, Einstein’s equations reduce to Newton’s in the limit as velocity goes to zero (and other conditions, depending on the version of relativity theory in question). Moreover, of course, the equations of motion in special relativity are derived from the classical (Newtonian) equations by applying the Lorentz transformations (and the equations of general relativity are derived in turn by applying yet more radical transformations). This would seem to imply that Newton’s laws captured something important about the relational structure of motion, which is evident at velocities that are low relative to the speed of light (the context in which they were formulated), and which survives in the newer theory. Thus, Newton might have been wrong about the constitution of the world—that it consists of absolute space and time in which there is gravitational action at a distance and events that are simultaneous for one observer are simultaneous for all observers—but right about an important aspect of its relational structure.
A different example that Worrall gives concerns the history of optics. Light was conceived in the 18th century to consist of particles. This theory accurately accounted for simple cases of reflection and refraction and for the dispersion of light in a prism. But this theory was replaced in the early 19th century by the very different conception of light as waves in a mechanical medium. The new theory generalized the predictions of the reflection and refraction of light dramatically. It was in turn replaced in the mid-19th century by Maxwell’s theory of light as waves in a non-mechanical, ontologically new entity, the electromagnetic field. This theory in turn was replaced in the early 20th century by the theory that light consists of photons—so, we’re back to particles, only particles now that obey wave equations. These are all quite different ontologies of light—different descriptions of what light is—but at each stage, the new theory preserved in at least an approximate way the equations of the old theory. Worrall describes in detail how, for the transition from the mechanical to the electromagnetic wave theories (Fresnel to Maxwell), the reflection and refraction equations were exactly preserved, despite the ontological change.
Thus, according to this form of structural realism, we should not have too much confidence that the ontology that is implied by our best scientific theories will stand the test of time. But that is compatible with having great confidence that the laws of those theories will do so, at least approximately. And note that the notion of approximation here, although left vague, should be understood to include the continued predictive accuracy that the laws enjoyed prior to theory change. Newton’s laws didn’t become less accurate than they ever were after the development of relativity theory, and neither did Fresnel’s after the development of Maxwell’s equations. Scientific progress, although perhaps not ontologically cumulative, does seem to be predictively and structurally cumulative. That is (my view of) structural realism in the philosophy of science.
In this light, it is clear in what way Carroll’s claim that “the laws underlying the physics of everyday life are completely understood” is correct. Quantum field theory, general relativity, and the Standard Model provide predictions of the most basic phenomena to an astonishing level of accuracy. In a commonly cited example, the magnetic moment of the electron (the amount of torque it undergoes in a magnetic field) is predicted by quantum electrodynamics to an accuracy of one part in 106 (Veltman 2018, 261–263). Perhaps more to Carroll’s point, the predictions have been thoroughly tested and found to be accurate in particle accelerators, where conditions such as energy and velocity go far beyond what Carroll regards as “everyday life.” Although the physics community very much yearns to break the Standard Model somehow and thus find “new physics,” I believe nobody at this point thinks that’s going to happen at the levels of energy we have so far been able to employ in experiments. Since these levels already exceed those of everyday life, it is fair to infer that the body of physical law we have arrived at is not going to change for the everyday regime. Ever.
So for physics, is it “the end of history”? Have we shown that there will be no more revolutions? Not according to structural realism. A main point of structural realism in the philosophy of science is that there can still be revolutions with their implied changes of ontology that do not (or not much) affect the status of the laws of previous theories in the regimes in which they are predictively adequate. It is conceivable that the present body of physical law will be swept up into a “Grand Unification” or “Theory of Everything” or, more radically, some entirely new physical theory that changes our conception of the constitution of nature radically. But, if structural realism is right, then such a new theory will very probably leave the laws of quantum mechanics and general relativity largely intact as limiting cases of the new theory, where the “limits” in question occur in the everyday regime.
Is this the right interpretation of Carroll? The following passage might raise doubts:
What there won’t be is some dramatic paradigm shift that says “Oops, sorry about those electrons and protons and neutrons, we found that they don’t really exist. Now it’s zylbots all the way down.” … The view of electrons and protons and neutrons interacting through the Standard Model and gravity will stay with us forever — added to and better understood, but never replaced or drastically modified.
This might seem to deny the possibility of a new theory with radically different ontological implications. But I don’t think that is what Carroll is saying. And the reason is that I don’t think Carroll regards “electrons and protons and neutrons” with much ontological seriousness even today, without the disruption of any dramatic paradigm shift. Particle physics today—and this is definitely Carroll’s own view (see here)—regards particles as “field quanta,” meaning energetic regions of fields (e.g., the electron field, the electromagnetic field, etc.) that show up in observable interactions as particles. In this view, what is truly fundamental are fields, and what they consist of—if they consist of something still more fundamental—is anyone’s guess. In the meantime, fields in quantum field theory are a part of the mathematical formalism of the theory. Carroll’s point, I think, is that the discovery of something more fundamental is not going to change (“drastically modify”) that formalism. If a dramatic paradigm shift leaves quantum field theory with the status of a limiting case that is largely valid in the everyday regime, as structural realism (and Sean Carroll, if I have him right) says it will, then there will still be fields and field quanta, i.e., electrons and protons and neutrons.
Could the bounds of the everyday regime change? I have spoken of the energies in particle accelerators exceeding those of the everyday regime, but couldn’t we eventually reach the point where we utilize such energies on an everyday basis? Yes, it seems conceivable that we could, and, since civilization depends on the exploitation of energy, I hope for the sake of humanity that we do. But I don’t think that would invalidate Carroll’s claim. Clearly, he is talking about “everyday life” as we live it now. Come the day when everyday life extends beyond the realm where current physics is valid, I imagine Carroll will be happy to be “refuted.”
Rather than worry about what everyday life might entail a thousand years in the future, a better way to look at the remarkable achievement Carroll is pointing to is to think about what everyday life has meant for humanity throughout the whole of our existence up until now. The basic laws that underlie the physics of all the phenomena of everyday life—all of it—for the whole of human history, have now been discovered. That is something we have to thank the 20th century for.