The Causal Theory of Views

The Causal Theory of Views

Lee Smolin [12.19.19]

An event has a view of the world. First, let me tell you what I mean by a view. A view is the information that that event has about how it fits into the rest of the world. That includes who its parents were (by which I mean the events in its past that gave rise to it) and how much energy and momentum was propagated to it from them. If I am an event, my view of the world is what I see when I look around. I see light comes to me from the past, which I perceive as a pattern of colors, which come from photons of different energies striking my eye. That's my view; it's a property of a moment. That contains all that I, as an event, know about how I fit into the rest of the world.

Now, if you know the things that I just said were real—the events, the causal relations, the distribution of energy and momentum flowing—I can tell you what the view of each event is, but I can also flip it around. There's a dual description in which I just say what the views are and that's the whole description. So, I just say there's a view, and that view is that makes a kind of picture. You see the sky, a two-dimensional sphere around you, and there are some colors, which are photons coming in of different energies—that's the view. I can hypothesize that all that exists in the world is views and a process that continually makes new views out of old views. That's what I call the causal theory of views.

LEE SMOLIN, a theoretical physicist, is a founding and senior faculty member at the Perimeter Institute for Theoretical Physics in Canada. His main contributions have been so far to the quantum theory of gravity, to which he has been a co-inventor and major contributor to two major directions, loop quantum gravity and deformed special relativity. He is the author, most recently, of Einstein's Unfinished RevolutionLee Smolin's Edge Bio Page


[ NOTE: The text of the interview has been edited for clarity and does not correspond exactly to the audio or video.]

In my life as a theoretical physicist, I've worked on several different problems, but primarily on quantum gravity—how to unify or bring together quantum theory and quantum physics with space, and time, and gravity, and relativity. I've also given a substantial amount of time to the problems in the foundations of quantum mechanics, to making sense of quantum mechanics. In the last five or so years, my work on these two different problems has converged. Now, everything that I'm working on, all the progress that I'm making, is on a theory that attacks and, in my opinion, resolves both the problems of quantum gravity and the problem of making sense of quantum mechanics.

There are two ways that people study the measurement problem and the other problems in the foundations of quantum theory. There are a lot of people who take quantum mechanics as given, that is, as developed and put in final form in the late 1920s by Werner Heisenberg and his collaborators. They note that we're uncomfortable with it, that the theory has certain puzzles, has behavior that is not very intuitive. But they presume the problem is not in nature but in us, in our understanding and search for different ways to express the physics or the principles of that theory without challenging or changing that theory. Unfortunately, that's most of the work done in what's called quantum foundations.

Then there's a small number, and I'm one of them, that take the problems in quantum mechanics as evidence that the theory is wrong, or at least incomplete. This was Einstein's view, de Broglie's view, Schrödinger's view. Among contemporaries, it's Roger Penrose's view. We hold that the reason the theory is hard to understand and has puzzles is because it's the wrong theory, and we look for a completion of the theory. What we mean by that is a theory that would do something that quantum mechanics does not do, which is give a complete description and explanation for what happens in each and every individual process. That's what I've been trying to do for a long time. Since I was in graduate school, every few years I write a paper or two papers about that.

There is a key idea that underlies my work in that area, and that's the problem of what's called nonlocality. This is a very important hint. Nonlocality is the phenomenon—also sometimes called entanglement—that if you have two quantum systems and they interact and then separate, they share properties, in that the choice of what property we measure on one of those particles, even if they're very far away from each other, affects what's measurable in the other particle. That's a statement of what's sometimes called Bell's theorem, and it's been put in a form where it can be tested experimentally. The experiments clearly show that the assumption that the two particles are independent because they're far away from each other is wrong. If you want a complete description of the results of experiments on these kinds of systems, and not a statistical description in terms of probabilities, which is all that quantum theory gives in these situations, you have to posit explicit interactions and communication between the two particles.

I believe that's right. That's what would be called a nonlocal theory. Traditionally, we call it a nonlocal hidden variable theory. There are a number of these—Louis de Broglie invented the first, called pilot-wave theory. It was laughed out of town by people in the late 1920s. There were theorems that showed it was impossible in the early thirties, and it was basically dropped by de Broglie and everybody else and then rediscovered by David Bohm. That's the first nonlocal hidden variable theory we have, and there are others. I like the de Broglie-Bohm pilot-wave theory as an example, which shows that route is possible, but I don't believe the assumptions of that theory, and I don't think it's a route to a deeper understanding of nature, so we won't talk about it.

The theory that I've been looking for would take advantage of the fact that the notion of locality and nonlocality is key to understanding quantum mechanics, and then try to understand that with the lens of the unification of quantum physics with space and time, which is quantum gravity. In both approaches, there's a principle, which is the idea of relational physics—that the degrees of freedom, the properties of whatever it is that's dynamical that you're studying, arises from dynamical relationships with other degrees of freedom. In other words, you don't have absolute space, you don't have particles that occupy points or follow paths or trajectories in absolute space. You have many particles which, between them, allow you to define relative motion. This principle was introduced by Leibniz and developed by Mach, and it was a very important inspiration for Einstein in general relativity.

I have tried to use that principle in making nonlocal hidden variable theories. Through the commonality of having relational descriptions in quantum theories of gravity and in hidden variable theories or completions of quantum mechanics, I'm able to invent theories that address both problems of quantum mechanics and quantum gravity. Particularly in the last eight years or so, there is a version of these theories that is developing, which I'm very excited about.

The theory that I'd like to talk about first is an approach to both quantum gravity and quantum foundations that we invented and developed with Marina Cortes, called energetic causal sets. This is a theory that is sometimes called a realist theory. A realist theory is an approach to quantum mechanics that describes what's there. It has nothing to do with measurement, or what we think, or our consciousness, or our knowledge; it is just a description of nature as it is. A realist theory has "not observables," as quantum people talk about it, which may or may not have definite values—but what John Bell called "beables." A beable is something that always has a definite value. The idea is to reduce and get rid of the confusing use of observables, which are put forward in a very nonrealist or operational point of view, and replace them with talks about beables. When you're talking about beables, you're talking about the ontology, what the theory says is fundamental and real. So, let me tell you what is fundamental and real in this theory.

First of all, events. Events are things that happen. Events create other events. In this theory, there are always new events being made. An event has parents, if you will, which give rise to that event. The parents might shoot out an electron each, and the electrons combine—that's the event. The parents are the two events in which the two atoms shot out those two electrons. We have by that a causal relationship—the intersection of the two electrons is caused by the atoms that emitted the two electrons. These causal relations are real, the events are real, and the process that builds the new events or creates the new events continually from the old events is real, and the causal relations are real. Space and time, in the sense of spacetime, in the sense of a preexisting geometry with points and coordinates, is not real, is not fundamental, is not part of this description. Something else that is real and fundamental is energy and momentum. That's why we call these energetic causal sets, so that these causal relations carry energy and momentum, and the events inherit that energy and momentum, and then split it up and send it off.

What's emergent, what is not fundamental but emerges at a higher level the way that temperature and pressure emerge, is spacetime. The idea of elementary particles propagating classically and the idea of the quantum state and the quantum description are all emergent from this fundamental ontology.

There are many results that we've achieved from these theories, but I want to talk about one aspect of it. I've been thinking about how to implement the idea I'm about to describe for all my career, and I finally was able to do it. It's the following idea. An event has a view of the world. First, let me tell you what I mean by a view. A view is the information about how it fits into the rest of the world that that event has. That includes who its parents are and what the energy and momentum was that was propagated to it from them. My view of the world is, I look out and light comes up from the past and I see a pattern of colors, which come from photons of different energies striking my eye. That's my view; it's a property of a moment. That contains all that I, as an event, know about how I fit into the rest of the world.

Now, if you know the things that I just said were real—the events, the causal relations, the distribution of energy and momentum flowing—I can tell you what the view of each event is, but I can also flip it around. There's a dual description in which I just say what the views are and that's the whole description. So, I just say there's a view, and that view is and hears a kind of picture. You see the sky, a two-dimensional sphere around you, and there are some colors, which are photons coming in of different energies—that's the view. I can hypothesize that all that exists in the world is views and a process that continually makes new views out of old views. That's what I call the causal theory of views.

I take this theory very seriously, that is, this is not just talk; there's mathematics for this, it's well worked out, and we've developed various consequences of it. The universe is a collection of partial views, because each of these is just part of the universe of itself. That's all the universe is, fundamentally, in this story. 

One consequence of this is that quantum mechanics cannot be fundamental. One of the principles of that theory—called the uncertainty principle—says that you cannot know, at any given time, the values of both the position and the momentum of a particle. But in the causal theory of views, space and time are not fundamental—they are nowhere to be found in the fundamental description. Momentum and energy are there, but, because there is no space, there are no positions, and we cannot even state the uncertainty principle.

At a higher level of organization, we recover quantum mechanics as a description of an emergent phenomena, namely motion in space and time. At that level, space is emergent, and so is the uncertainty principle. So also are the notion of locality and the distinction between local and nonlocal phenomena.

What are we able to show from this? We can show that spacetime emerges in a way in which you give an approximate description of all these causal relations between the views, and the views are represented in spacetime by the light cone, that is, an event and the light cone of photons going up to it. This is very important because there are a number of different approaches that start off saying that space and time are discrete and they can't really recover the world of spacetime as it is. This was one of our first important results. Then, at the level of description where space emerges, you also get quantum physics, that is, coordinates in that space, measuring those interferes with measuring those momenta. So, quantum mechanics emerges.

There's an interesting fact about that, which was one of the most exciting moments I've had in all my research. Normally, physics describes systems from the outside, and when we talk about interactions or forces, we talk about locality. Locality means that things interact with other things in the same place, or there's a force, like gravity, that falls off like a square in the distance. It depends on the notion of distance. But we have no space in this fundamental theory. So how can we formulate dynamics? The way we can formulate dynamics is we measure the difference between the views. We can pick two events and look at how different or how similar their views of the rest of the world are, and we can define an energy of the whole system, a kind of potential energy of the system, which is a function only of the differences of the views.

Julian Barbour and I worked together in the late eighties, and we invented a measure of the differences of the views in such a system, which we called "the variety." The variety measures the total amount of differentiation or indeed variety in a system of relations. You look at all the pairs of views and measure how different they are, and then you add that up over all the pairs. I put this into the theory in order to give some dynamics and some interactions. Lo and behold, what I get out when I make the right approximations is the key ingredient of the de Broglie-Bohm hidden variable theory. So, I recovered the de Broglie-Bohm hidden variable theory from this theory. I would have to talk technically to explain how that is. That force of the variety is exactly the right interaction to put into a system of views to so that, when you specialize the views to be views of some particles moving slowly and nonrelativistically, the force becomes the same as is found in de Broglie-Bohm theory, which makes it equivalent to quantum mechanics. That's what I mean when I say that quantum mechanics is recovered.

There's a version of this theory that we developed with Wolfgang Wieland, which is a spin foam theory. A spin foam theory is a version of loop quantum gravity that we developed with Carlo Rovelli and Abhay Ashtekar. We made contact with that world, and we can discuss loop quantum gravity in a way that has events and causal relations. That is part of everything that's coming together in the theory.

Now that I've laid out some of the aspects of energetic causal sets and the causal theory of views, let me situate them with respect to research in quantum gravity. Presently, we have six or eight approaches to quantum gravity that get partway there. The two that are most developed by quite a bit are string theory and loop quantum gravity. I was one of the inventors of loop quantum gravity. I'm impressed by it, I care about it, but there are problems that it doesn't solve. There are places, in my view, that we've been stuck for a while. The same thing is true about string theory. I'm excited about string theory, but as a fundamental theory, it has reached roadblocks and stumbling points that we haven't figured out how to get around.

This is an approach that, as I said, incorporates or can incorporate some elements of loop quantum gravity. We haven't worked out very much. It could maybe make contact with string theory. It's new and it certainly doesn't have the mass amount of work, and results, and people involved that these two theories have had. It's at the beginning. There are also four or five small programs in quantum gravity: causal sets itself without the energetics, causal dynamical triangulations—the word causal comes up a lot—asymptotic safety, and several others that I can name where the results are very tantalizing, but there are also some problems that don't go away.

We're trying to incorporate elements of previous ideas, but it's a different start. It's a fresh start. It's what we should be doing, rather than sitting on our little hills that we've built up and throwing stones at each other, if I can make a metaphor. We should all be meeting down in the valley and trying to start afresh, and this is an attempt to do that. For me that meant starting with principles. And the principles I mentioned about relational theories and wanting a realist theory of quantum mechanics are fundamental principles that motivate the work to try and make a fresh start. There are a few people who are contributing to this, but not many. It's at the beginning. I feel incredibly excited about this because of the way it brings together the different strands of my work the last thirty-something years.

This work builds on all the thinking and work I did, arguing for the fundamental nature of time. It brought my work on that question on time together with Marina Cortes' work on the importance of irreversibility, the fact that fundamentally the universe is irreversible. It brings together my interest in the idea that the laws of nature evolve, which led to the idea of the landscape of theories and to my idea of cosmological natural selection as a way to explain the choice of the laws, including the values of the different coupling constants which are the parameters in of the standard model of physics.

The idea that string theory has a problem, and that there are a large number of versions of string theory and they all lead to different predictions for the dimensionality of space in the elementary particles wasn't new in 2003. It was clear to a number of people already by 1987, 1988. Among them was Andrew Strominger, who's a good friend and now a professor at Harvard. At first there were five different string theories. Then there were these Calabi-Yau compactifications in 1986, of which there were hundreds of thousands. Compactification refers to shrinking down and folding up the six extra dimensions. They're much smaller than the three dimensions of space that we observe. Andy discovered in 1987 that there were a vast, probably uncountable number of versions of string theory, so it was much worse than the hundreds of thousands even from Calabi-Yau compactifications. He published this in a theory called torsion in string theory. He came to me with this problem. He was very upset that there were so many versions of string theory we wouldn't be able to explain anything, because whatever the experimentalists see, we will have a version of string theory that agrees with that.

At that time, I was just recreationally reading Lynn Margulis, Richard Dawkins, Steve Gould, and all that literature. I thought, here in biology we have the process of natural selection, which chooses parameters, namely the genes, in a way that leads to a biosphere that has a lot of structure and complexity. And here we have the universe evolving, which somehow chooses laws and parameters of laws in a way that leads to a universe that is full of complexity and life. I thought, can I take the methods and the ideas from population biology and evolutionary theory and apply them to physics?

In biology, they talk about the landscape. It's an enormous space of all the possible gene sequences that could be the DNA of a species, and they have a function on that landscape, which is the fitness of a being with those genes. They argue that a population will tend to cluster at the top of hills of that function, where there's the most fitness. I took all of that, with the language, and found a scenario to realize it in cosmology. The details of the scenario don't matter, but the key hypotheses were that black hole singularities give rise to new universes and that each new universe had slightly different parameters of the laws of physics. The more black holes a universe produced, the more universes and, hence the higher the fitness. I assumed the changes of the parameters were small in each generation so that you accumulated fitness. I published this as a paper in 1992 and this was the subject of first book I wrote, Life of the Cosmos.

This simple scenario implied two falsifiable predictions. First the heaviest neutron stars were less than twice the mass of the sun, second that if cosmological inflation occurred, it was a certain kind called single parameter-single field. 

Both held up well, until very recently, when I understand they are each challenged, but not definitively falsified, by recent observations.

Later, in 2003 a group at Stanford discovered a way to solve another problem for string theory in a way that produced a description of a landscape of a vast number of string theories.

However, their use of the landscape was rather different. Rather than using the analogy to natural selection, as I had, the Stanford group used the idea of the anthropic principle, which people like Martin Rees and colleagues had developed as an explanation for the parameters of cosmology and physics. My main problem with this version is that it makes no falsifable predictions. 

I had the privilege of being part of the invention of loop quantum gravity. Each of us, Abhay Ashtekar, Carlo Rovelli, and myself, put in an essential element. I'm very proud of that. Whether that theory is right or not, it did take off and become something that hundreds of people work on. There are conferences. You can measure progress. I'm proud of that. I'm not sure it's right, but it gives a detailed and, in many ways, compelling, of one way to solve the problem of quantum gravity. That is, it describes a way the world might be-and that isn’t nothing. I think Carlo and Abhay would say the same thing. I've strayed more outside of the community we created then they have. They have chosen to emphasize developing this one theory, while I have preferred to continue to invent. What I'm good at is initiating and inventing. I'm not good at leading a large group of people and trying to get them to march in the same direction. They do that well.

When I was invited to be one of the first founding scientists at Perimeter Institute, the founding donor, Mike Lazaridis, and the founding director, Howard Burton, explained to me the principles that the new institute was to be based on. This institute was to be different and better. One of them was that we were going to discourage silos. You couldn't come and just work on one thing. There was the principle that everybody is better off with some opposition. So, if we're going to have people working on quantum gravity, we're going to get loop quantum gravity people, and string people, and twister people, and everybody's going to be uncomfortable. There would be no defaults, no entitlements. And we've kept to that principle. We have people studying quantum gravity from a variety of different points of view, and this gives our research center and our research area a different flavor from others.

One way to measure how well we've done is that in the area of quantum gravity, which is not a big area, but it's grown as a result of our work, something like half of the faculty who were hired into faculty positions in the field the last nineteen years since we started are our postdocs. We've been like a school, and indeed the central school, and we have communicated this ethic of tolerance and disagreement. It's been a good thing.

Let's put on the table the question that I and other people are trying to address when we talk about consciousness, or qualia, as it's sometimes called. This is what David Chalmers calls the hard problem. It's been there since Galileo, who spoke about it, Leibniz spoke about it. Let's say I could completely describe what's going on in your neurons and all the chemicals and fields. I would not have any reason to deduce that you have sensations and have the experience, which is only open to you, of seeing colors, and hearing sounds, and thinking thoughts, and feeling emotions and pains. Leibniz said that if the brain were a mill and you go inside the mill and see all the parts working, you would never deduce that the brain was associated with a mind that had an experience that could say, "I experience." There is an explanatory gap.

There are stories that people tell. There's Mary, the colorblind neuroscientist. Mary is colorblind, as the title says, and she is the world's greatest neuroscientist (this is 500 years in the future), and she knows everything there is to know about how the brain processes vision and color vision. One day, somebody invents an operation that can give her back color-sight, so she undergoes the operation. She wakes up the next day, takes off her bandages, goes outside, and she looks up at the sky and says, "I'm the world's greatest neuroscientist. I knew everything about color vision, but until now I didn't know what the color blue was like, what it was like to have the sensation of the color blue." Now, the philosopher who invented that argument asks us, is Mary right? Is there something that she now knows that she didn't know before?

If you think the answer is yes, then that's the question that is to be addressed. People who don't think there is a question deny that she actually knows something that she didn't know before. That's one of the different ways that people talk about the question. Another way is the conceivability of zombies argument. Given everything we know about science, it's perfectly possible that there are human beings who don't have any inner experiences, who, because physics, and biology, and chemistry explain everything that happens in their brain, explain what they hear, what they see, what they think, what they decided to do. It might be that I know I have these inner experiences, but it might be that I'm the only one, or my best friend is a zombie and doesn't have any experience going on. The fact that this is conceivable, David Chalmers argues, means there's a gap in our knowledge, because if we understood what Mary knows now, we would realize that zombies were impossible, that part of being a human being is to have these inner experiences.

Probably the sensible view, and my view for most of my life, has been that this is an interesting problem. It's a scientific problem. I hope someday in the progress of science it's addressed, but it's too hard to work on now. There was only one view of the many views that are offered that seemed to make sense to me, which was invented by Bertrand Russell, and this is called Russellian neutral monism, and it was endorsed by Arthur Eddington and a number of other people at the time. That's roughly that when we describe atoms and elementary particles and forces, we're describing, especially in a relational universe, how things appear to evolve from the point of view of things outside them. But atoms and electric fields might have some inner essence. There might be something about what it's like to be an atom, or molecule, or a cell, or a bat. This is another aspect of being material. This is not against materialism, or physicalism; this is saying there's another aspect that we only know of because we experience directly the chemical and electrical events in our brains.

That has seemed to me a reasonable point of view. I wrote a paper about it in college. It seemed to me then as far as you could go. Up until very recently, that was my view. To leave you with something tantalizing, recently, I see a way to address some aspects of that question within the physics that I've been developing.

There is a view that the world is fundamentally mathematical, which means that there is some mathematical object, a solution of equations or a system of equations that corresponds to the world in the sense that every property that's true about the world is a theorem that you can prove about this mathematical object. I don't think that's true. There are a very few things that are properties of the world that are not properties of any mathematical object. One of them is that it's always some time. There's the present moment in nature, whereas all the trueths of mathematical objects are timeless. There are logical implications, and logical implications can model causal relations, but they're not the same. Another one is the sensations of qualia, of what it's like to see red and blue.

Mathematics will play a role. Science will be able to incorporate some testable hypotheses about colors, and sounds, and sensations, and they will become part of the mathematical models that we make of the world. I also think that we'll never get back that, to me, very romantic but wrong idea that our job as theoretical physicists is to find an equation that represents all of reality.