Introduction

I am currently writing a series on a recent article by Sean Carroll. This will take me several posts to respond to it fully, so I am going one section at a time. My previous post gave an introduction to the topic, and covered Carroll's own introduction. In this post, I will discuss the first section of the article, entitled "What Does Why Mean?"

In this article, Carroll is focusing on the "fundamental question of metaphysics." If science and philosophy are largely concerned with questions of how things are, but the more fundamental question is why things are. But before we answer this question, we need to ask what we mean by Why, and what sort of answers we might expect. The purpose of this section is to address that issue.

Carroll's conclusion in this section is that there are five possible answers to this question, and I actually agree with him on this point, so I will come to it first. His five possibilities are:

1. Creation. There is something outside reality which brings it into existence, and/or sustains it. This is often identified with God or some sort of necessary being.
2. Metaverse. Reality is just one member of a wider series of structures. For example, one might have pocket universes emerging from a larger structure, or a bouncing universe where one universe dies and then another begins, stretching back to infinity, although Carroll intends his concept to be more profound than this. This wider structure will be similar to our own reality (for example governed by similar rules, so it is distinct from the idea of a creator), but larger and at a more fundamental level.
3. Principle. Reality is the way it is because of some underlying principle, perhaps simplicity or beauty.
4. Coherence. The concept of "nothingness" is incoherent, so anything except a universe consisting of something is non-viable.
5. Brute Fact. Reality is the way it is, and that's it. Eventually, the chain of explanation will reach a point where we can't go any further. The is no underlying reason why that terminus has to be what it is -- it could be different -- but it just happens to be that way, and we have to live with it.

My main quibble with all of this is Carroll's use of the word reality here as a synonym for the universe. His definition of the universe was:

I'm using "universe" here to refer to the entirety of physical reality. No judgement is implied about whether things other than physical reality can be usefully said to exist.

If so, there is a distinction we need to make between physical reality and reality. The two can only be equated if physical reality is all that there is. However, this is something that would be disputed by those who hold to a creator, Professor Carroll's first option. These people would claim that God is part of reality, but not physical reality. This issue be avoided by replacing reality with physical reality in all of Carroll's five possible answers above.

The second thing I want to note immediately is that there is a great deal of similarity between the Creator and Brute Fact options. Both of them postulate that the chain of causality terminates at some point. The difference is in what they believe about the nature of that point of termination. Those who advocate a Creator will say that He (or It) can only be the way He (or It) is. The classic example of this is the being of pure actuality in the Thomist tradition. The Potentia list the possible states into which a being can change into, similar to how an electron orbiting an atom has different energy levels or eigenstates of the Hamiltonian. These energy levels would represent the different possible potentia. In a classical Aristotelian metaphysics, only one of these can be occupied at a given time, and that state exists actually. The remaining states can be said to exist potentially, since there is a possibility that the electron will be excited into one of them, or decay into a lower energy state. God, being a being of pure actuality, has no potentia. He cannot be other than as He is (and is incapable of change). Thus while it is necessary and possible to explain why the electron is in that particular eigenstate rather than another, we cannot sensibly ask why God is as He is because there are no other options.

In these terms, the brute fact approach states that the sequence of explanations ends with something that is a mixture of potentiality and actuality. It could be something else, but happens not to be, and there is no underlying reason why it is in the state it is. Of course, this precise terminology can only directly be applied if the sequence of explanations ends with some sort of being, rather than (say) some abstract principle such as a law of quantum gravity. But even in the case of the abstract principle, we can still invoke a similar principle. There are numerous different ways the law of gravity could exhibit itself (for example, with different values of the ratio between the Planck mass and the Higgs Mass); one of them is actual, and the others are potential.

A second quibble is that Proffessor Carroll doesn't mention a sixth possibility, namely an infinite regression of explanations. Presumably that means that he believes the arguments against this are sufficiently strong that the possibility is not worth mentioning. Since I agree with him that the arguments against this view are strong, I won't discuss it further. But some will be disappointed to see its omission.

Mechanism and Purpose

So having started with Carroll's conclusion, and broadly accepted it, I ought to go back to the start of this section and point out where I have issues as he guides us to that conclusion. The first one arises in his very first paragraph. Since the time of Aristotle, philosophers have tried to understand what the possible senses there could be for why something is.

For our limited purposes here, it should suffice to distinguish between how the universe came to be and what mechanism (if any) might have brought it into being – corresponding roughly to Aristotle’s efficient cause – and the reason why (if any) it exists – corresponding to the final cause. Aristotle conceived of final causes teleologically, as ends or purposes. Here we’re being a little broader, expanding the category to include anything that would qualify as a "reason why." Let us label these notions mechanisms and reasons for short.

So Carroll manages to get both the efficient cause and the final cause wrong in two consecutive sentences. Firstly, the efficient cause is not a mechanism. It is a substance. In a weak interaction event, an electron decays into a W^- Boson and anti- neutrino. The mechanism is whatever it is that lies behind the weak interaction. The efficient cause of both the Boson and the anti-neutrino is the initial electron. And this is always the case: an efficient cause is always a being, such as an electron, quark, or protein molecule. This is a key difference between classical theories of causality and modern theories of causality, which do look for a mechanism or event behind the change.

Next we have the final cause, which Carroll denotes as an end or purpose. End is not actually too unreasonable, although we have to qualify it a bit. The final cause once again points to another physical state or potentia. So, among the final causes of the electron are the W^- Boson and anti- neutrino (with the particular energy, momentum and spin from that particular decay). So we can correctly relate the concept with an end, but only if we mean the right thing by end. Equating final causality with purpose has a long history, so Professor Carroll shouldn't perhaps be criticised too much for making the same mistake. The definition was, I think, first proposed by the Stoics including Cicero, and entered Western European thought during the Renaissance. So if you are discussing post-Renaissance thought, then this definition is reasonable. But Carroll is referencing Aristotle here, and (no doubt somebody will correct me), I believe that among the major Aristotelian philosophers, this definition of final causality was rarely used. Of course, Aquinas argued that because of final causality, there had to be a guiding purpose behind matter (which he equated with God), but that is arguing from the Aristotelian principle to the Will of God, not equating final causality with purpose. He could hardly have been arguing "Because there is purpose in nature, there has to be a purpose guiding it," since that is clearly circular.

Finally, after all his good work saying that there are different ways in which one could answer the question "Why," Carroll decides to lump them all together under his own definition of final causality. Oh well.

In my previous post, I spent some time describing the difference between deism and theism. And this is one example of why I had to do that. Mechanisms and reasons are categories that are used in the modern philosophy of physics, which underlies deism. They are not categories that a classical philosopher or classical theologian would find useful, at least not in the sense that Carroll means them. Of course, that leaves the question of whether modern philosophy or classical philosophy are best used to describe modern physics, but that is another topic. I discussed earlier how the Aristotelian concepts of efficient and final causality fit in with one quantum physical process, and we can also see how the classical notion of mechanism doesn't fit in so well.

Carroll is searching for the answer to the question of what mechanism brought the universe into existence, a scientific question that ought to be informed by the best of modern science. I agree. But then asking the reason is different. Aristotle believed that final causality is a fundamental feature of reality. Modern Physics disagrees. Rather than describing things in terms of effects and causes, it states that the universe is described by patterns, the laws of physics. These relate different conditions at different times and places to each other, "typically by differential equations." Every "Why" question is reducible to those laws and the initial conditions. The laws are a description of patterns, a Humean account of the laws of physics.

But of course, Aristotle's vision of final causality is not what Carroll described it as. And the disagreement was promoted when Galileo and Newton were King, and we now know that they were both fundamentally wrong about physics. So Carroll is comparing something that was not what Aristotle proposed against an incorrect physics. This doesn't really tell us much of use at all, either about Aristotle or about physics or about reality. Crucially, contemporary physics no longer makes use of differential equations in the same sense that Newton did. Newton's laws of motion relates the location and time of the particles in the universe, and their interactions, with a differential equation. Plug in the initial conditions, solve the equations, and you can work out what will happen in the future. But that was disposed with a century ago.

Quantum mechanics moves a step away from this. Here we have a single particle wavefunction which evolves according to a differential equation, except when we take measurements. What this wavefunction represents is an unsolved philosophical problem, but not one that need concern us here. The evolution of the wavefunction is deterministic, but that is not the whole story, because it only describes the probability amplitude that a particle is in a given state, not what is actually happening with the particle (if the question of something actually happening makes sense in quantum mechanics). But the process of solving problems in quantum mechanics is similar to that in classical mechanics: it involves solving a lot of differential equations.

But, when we get to quantum field theory, the merger between special relativity and quantum physics, the evolution of matter is no longer described by a differential equation. Instead we have integrals over functions or operators. The Feynman diagram expansion (while only a guide, since it describes unphysical unrenormalised particles) is a good picture to have in mind. We have particles decaying into other particles, particles absorbing other particles. Each coherent route between an initial and final state is possible. Each one carries a certain amplitude. To get the final probability for a particular final state, we need to add up each of these amplitudes, and square it. What this means is that the old view of physical laws, as outlined by Professor Carroll, is no longer accurate. We are left with something far closer to Aristotle's vision of causes and effects. The difference is that we are now able to calculate the amplitude for each decay channel, while Aristotle could only list them. This is significant because metaphysics should meet with the most fundamental physics, not approximations to it. We have no reason to suspect that a theory unifying gravity with the standard model will be any different in this regard from the standard model.

But Professor Carroll's problems are deeper than this. He both wants to use the laws as an answer to the "why" question, and to have them as Humean descriptions of patterns. But that won't work, because a description of a pattern doesn't explain anything. It is a bit like asking "Why did that apple fall from the tree?" "Because of the law of gravity." "What is the law of gravity?" "It is a description of the observation that every apple falls after detaching from a tree." That is just circular. It hasn't explained anything. Nor does it give us any confidence that such laws work or would continue to work, because they are not linked to nature.

I would also agree that the laws of physics are a description, but a description of something that is inherent within nature. This could be a description of the ways in which particles interact with each other (the final causes), or, at a deeper level, a description of how God sustains the universe in accordance with its created nature. The laws of physics describe interactions between different beings. Therefore they cannot be fundamental; the notion of the beings themselves has to come first (because you can't describe an interaction before you define what it is that interacts), and the laws are either contained within that notion or follow from it.

It is certainly true that fundamental physical law is mostly described in terms of symmetries of the Lagrangian. But we have to ask why those symmetries in particular are satisfied. A call to symmetry is only a description of something that constrains the mathematical form of the description of the interactions between two beings. As, indeed, is a reference to locality, which is just as important in modern physics as a source of conservation laws. In classical physics, the conservation of energy and momentum arise as a consequence of local symmetries of the Lagrangian. In quantum physics, they arise because the Lagrangian is local, and only contains interactions at the same point in space-time.

Cause and effect

Carroll's argument in this section is that cause and effect language is no longer part of fundamental physics (although it might emerge at higher levels of physics). He wants to use this to say that it is pointless to think about the origin of the universe in terms of a cause. However, the reason why language of causality was removed from physics research is that it is not something that is easily representable mathematically. That doesn't mean that it isn't there, but that it is something we have to induce from looking at the structure of physical processes. When we talk about causes we mean that one state is destroyed and another created, and there is a relation between those two acts. When I see the Hamiltonian of quantum field theory, constructed from various creation and annihilation operators, I see one physical state being annihilated and some others created at the same place in space and time. This is a cause and effect process. That physicists don't usually use the language of cause and effect to describe this doesn't change that.

I would agree with Carroll that it is wrong to apply the principle of causality to the level of the universe, but for a different reason. Efficient and final causality apply to individual beings, and the universe isn't an individual being. But that doesn't mean that the idea of efficient causality and final causality aren't relevant for the question of whether there is something rather than nothing. We still have a chain of efficient causes. That chain must either continue indefinitely or terminate with something. We can ask, if it terminates, then what is the nature of the being at the end of it, and whether all such chains of causality terminate with the same being.

While we don’t currently know the once-and-for-all laws of nature, nothing that we do currently understand about physics implies any necessary obstacle to thinking of the universe as a fully law-abiding, self-contained system. In this case, there would be no such thing as the "cause" or "reason why" the universe exists.

This statement, to my mind is a bit bold. Of course, one could argue that the beginning of the universe points to something outside. Or a cosmological argument based on efficient causality. One could also add that nothing implies that the universe is a fully law abiding, self-contained system. One has to ask why those laws are obeyed, and what it is about matter that makes it only decay in such a way that momentum or electric charge are conserved (and calling on symmetry and locality principles doesn't help: one still has to explain why the action has those symmetries and is local, and why an action or Hamiltonian operator is useful in understanding how matter evolves).

I will also mention the indeterminacy of physics as another reason for thinking that the universe is not self-contained. The interpretations of probability which I find most reasonable, and which best fit what we do in quantum physics, treat a probability in itself as a measure of uncertainty, and probability theory as an extension to logic which deals with things that are uncertain. Thus we only need to resort to probabilities if some facts aren't known. If everything was known, we would be certain. Since we are uncertain about some details, we have to use probability. Quantum theory is indeterminate, meaning that one cannot make precise predictions even with a complete knowledge of the present state of the universe. One can only express things in terms of probabilities (or, more specifically, probability amplitudes). This suggests that even with a complete knowledge of the universe, some facts are missing. Which suggests that the universe is not self-contained. Are arguments of this sort completely solid? Well, there are some assumptions I made that can be challenged. But thinking along these lines still casts doubt on Professor Carroll's assertion that there are no obstacles to thinking that the universe is self-contained. We have to carefully define what is meant by self-contained. Certainly the definition used in classical physics, which stated that every event had a full in-universe explanation, cannot be sustained, on any interpretation of the meaning of probability. So presumably, Professor Carroll means something else by the term. It would be nice to know what it is.

And that is before we run into fifth way type arguments (not the argument from design, but an argument from final causality). We have to ask why brute lumps of matter should have a tendency towards certain decay channels and not others. To turn to the laws of physics to answer this question isn't an answer: the laws describe what happens, but don't explain it. To express this in terms of the underlying symmetries is not answering the question, just rephrasing it. Yet the behaviour of matter seems to be wholly rational, which suggests that there is an explanation. But one that goes beyond the scope of physics, and thus physical reality.

One can also try to explain things by appealing to deeper principles. So, perhaps, once we come up with a complete understanding of physics, it might turn out to be the simplest possible universe, or the most symmetric, or so on. As Professor Carroll notes, however, we would still need to explain why that principle held.

So then we come to the idea which Professor Carroll attributes to Leibniz that there is explanatory regression which terminates with a necessary being. This is a point where no further explanation is possible, because whatever the terminating principle is, it couldn't be other than what it is. This has the advantage that it ties everything up neatly. It is the solution to the problem that most theists adopt (albeit that they might not express it in precisely these terms).

However, I should make one observation here: there is a difference between a necessary being and a logically necessary being. A necessary being is the opposite of a contingent being. A contingent being is one which could in principle have an explanation outside itself for its existence and its present state (for a physical system, which eigenstate of the Hamiltonian is actual). A necessary being is thus one which cannot have such an explanation outside itself even in principle. Usually this is taken to mean that it could not be in any state other than the one that it is currently in. The reason why it is in that particular state is that it can't by its nature be in a different one. The second possible definition is the logically necessary being. This is something whose non-existence in its current state would be a logical contradiction. The two definitions are not equivalent; one can describe necessary beings which aren't logically necessary beings. For example, Thomas Aquinas' conception of God is necessary but not logically necessary. It is not completely clear which of these definitions Carroll is using. In the section I am discussing in this post, he contrasts the idea of a necessary being against a brute fact, which implies the first definition (a brute fact is a contingent being which terminates a chain of explanations). Later on, in section 3, where he presents arguments against the idea of a necessary being, he seems to adopt the second definition. But I will get to that in due course.

Conclusion

So, I have reached the end of Carroll's first non-introductory section, where he discusses the meaning of the word Why and what possible answers we might give to the question why there is something rather than nothing. His final list of possible answers is something I can accept, and since that is mainly what he takes from this section, I'm vaguely with him so far. But he has also sowed the seeds of a few problems that will come back to haunt him later. He made some minor mistakes in his understanding of classical philosophy, and that has caused him to commit to one particular view of the philosophy of science. He describes things in terms of mechanisms and reasons. However, it is not clear that his philosophy of science is correct. It is not what a classical philosopher or classical theist would accept. Although popular today, and the philosophy that drove the Galilean revolution, his idea that physics reduces to laws which are described in terms of differential equations is now out of date. His assertion that the universe is self contained is somewhat begging the question, and there are reasons given the indetermancy of physics to suggest that it isn't so. In Quantum Field theory, it is somewhat different. But I will discuss that in more detail when I come to his section 3. He is also a bit sloppy with some of his key definitions. At one point, he equates the universe with physical reality, and the next with the whole of reality. At one point, he defines a necessary being by contrasting it against a brute fact, but later defines it as something that is logically necessary. So we are already starting to see a few issues arising in his argument.

Next time, we move onto the question of what we mean by something and nothing. It might be a few weeks before I can post again; I have a deadline looming, and am a bit behind schedule in preparing for it. But the post will come when I am able to.