[Lee Smolin:] What is space and what is time? This is what the problem of quantum gravity is about. In general relativity, Einstein gave us not only a theory of gravity but a theory of what space and time are — a theory that overthrew the previous Newtonian conception of space and time. The problem of quantum gravity is how to combine the understanding of space and time we have from relativity theory with the quantum theory, which also tells us something essential and deep about nature. If we can do this, we'll discover a single unified theory of physics that will apply to all phenomena, from the very smallest scales to the universe itself. This theory will, we're quite sure, require us to conceive of space and time in new ways that take us beyond even what relativity theory has taught us.
But, beyond even this, a quantum theory of gravity must be a theory of cosmology. As such, it must also tell us how to describe the whole universe from the point of view of observers who live in it — for by definition there are no observers outside the universe. This leads directly to the main issues we're now struggling with, because it seems very difficult to understand how quantum theory could be extended from a description of atoms and molecules to a theory of the whole universe. As Bohr and Heisenberg taught us, quantum theory seems to make sense only when it's understood to be the description of something small and isolated from its observer — the observer is outside of it. For this reason, the merging of quantum theory and relativity into a single theory must also affect our understanding of the quantum theory. More generally, to solve the problem of quantum gravity we'll have to invent a good answer to the question: How can we, as observers who live inside the universe, construct a complete and objective description of it?
Most of my work as a scientist has been directed to the problem of quantum gravity. I like working on this problem a great deal, especially as it's the only area of physics I know of where one is daily confronted by deep philosophical problems while engaged in the usual craft of a theoretical physicist, which is to make calculations to try to extract predictions about nature from our theoretical pictures. Also, I like the fact that one needs to know a lot of different things to think about this problem. For example, it's likely that quantum gravity may be relevant for understanding the observational data from astronomy, and it's also likely that the new theory we're trying to construct will make use of new mathematical ideas and structures that are only now being discovered. So although I've worked almost solely on this problem for almost twenty years, I've never been bored.
I have days in which I spend the morning working on a calculation, to check an idea I had the night before, and then I'll go to a lunch seminar, where I hear astronomers discuss the latest evidence for some crucial question, like how much dark matter there is. Then I spend the afternoon studying the paper of a friend who's a pure mathematician, after which I meet a philosopher for dinner and continue an argument we're having on the nature of time. And what's wonderful is the way that these different subjects, which until recently were disconnected from one another, often seem to illuminate one another. Of course, sometimes it's not so ideal; teaching and bureaucracy take up a lot of time — although in reasonable doses, I must say. I love teaching also. But there are really many days when I feel very fortunate and can't imagine that I'm being paid to live like this.
For the last eight years or so — really, it doesn't seem so long! — I've been working with several friends on a new approach to combining relativity and quantum theory. We call this approach "nonperturbative quantum gravity." It's enabling us to investigate the implications of combining general relativity and quantum theory more deeply and thoroughly than was possible before. We aren't yet finished, but we're making progress steadily, and recently we've got the theory well enough in hand that we've been able to extract some experimental predictions from it. Unfortunately, the predictions we've been able to make so far can't be tested, because they're about the geometry of space at scales twenty orders of magnitude smaller than an atomic nucleus. But this is further toward a solution to the problem than anyone has gotten before — and, I must say, further than I sometimes expected we'd be able to go in my lifetime.
In this work, we've been combining a very beautiful formulation of Einstein's general theory of relativity discovered by my friend Abhay Ashtekar with some ideas about how to construct a quantum theory of the geometry of space and time in which everything is described in terms of loops. That is, rather than describing the world by saying where each particle is, we describe it in terms of how loops are knotted and linked with one another. This approach to quantum theory was invented by another friend — Carlo Rovelli — and myself, and also by the very interesting Uruguayan physicist Rodolfo Gambini.
The main result of this work is that at the Planck scale, which is twenty powers of ten smaller than an atomic nucleus, space looks like a network or weave of discrete loops. In fact, these loops are something like the atoms out of which space is built. We're able to predict that — just as the possible energies an atom can have come in discrete units — when one probes the structure of space at this Planck scale, one finds that the possible values the area of a surface or the volume of some region can have also come in discrete units. What seems to be the smooth geometry of space at our scale is just the result of an enormous number of these elementary loops joined and woven together, as an apparently smooth piece of cloth is really made out of many individual threads.
Furthermore, what's wonderful about the loop picture is that it's entirely a picture in terms of relations. There's no preexisting geometry for space, no fixed reference points; everything is dynamic and relational. This is the way Einstein taught us we have to understand the geometry of space and time — as something relational and dynamic, not fixed or given a priori. Using this loop picture, we've been able to translate this idea into the quantum theory.
Indeed, for me the most important idea behind the developments of twentieth-century physics and cosmology is that things don't have intrinsic properties at the fundamental level; all properties are about relations between things. This idea is the basic idea behind Einstein's general theory of relativity, but it has a longer history; it goes back at least to the seventeenth-century philosopher Leibniz, who opposed Newton's ideas of space and time because Newton took space and time to exist absolutely, while Leibniz wanted to understand them as arising only as aspects of the relations among things. For me, this fight between those who want the world to be made out of absolute entities and those who want it to be made only out of relations is a key theme in the story of the development of modern physics. Moreover, I'm partial. I think Leibniz and the relationalists were right, and that what's happening now in science can be understood as their triumph.
Indeed, in the last few years, I've also realized that the relational point of view can inspire ideas about other problems in physics and astronomy. These include the basic problem in elementary particle physics, which is accounting for all the masses and charges of the fundamental particles. I've come to believe that this problem is connected as well to two other basic questions that people have been wondering about for many years. The first of these is: Why are the laws of physics and the conditions of the universe special in ways that make the universe hospitable for the existence of living things? Closely related to this is the second question: Why, so long after it was formed, is the universe so full of structures? Beyond even the question of life, it's a remarkable fact that our universe seems, rather than having come to a uniform and boring state of thermal equilibrium, to have evolved to a state in which it's full of structure and complexity on virtually every scale, from the subnuclear to the cosmological.
The picture that emerges from both relativity and quantum theory is of a world conceived as a network of relations. Newton's hierarchical picture, in which atoms with fixed and absolute properties move against a fixed background of absolute space and time, is quite dead. This doesn't mean that atomism or reductionism are wrong, but it means that they must be understood in a more subtle and beautiful way than before. Quantum gravity, as far as we can tell, goes even further in this direction, as our description of the geometry of spacetime as woven together from loops and knots is a beautiful mathematical expression of the idea that the properties of any one part of the world are determined by its relationships and entanglement with the rest of the world.
As we began to develop this picture, I also began to wonder whether the basic philosophy behind it might extend to other aspects of nature, beyond just the description of space and time. More precisely, I began to wonder whether the world as a whole might be understood in a way that was more interrelated and relational than in the usual picture, in which everything is determined by fixed laws of nature. We usually imagine that the laws of nature are fixed, once and for all, by some absolute mathematical principle, and that they govern what goes on by acting at the level of the smallest and most fundamental particles. There are good reasons why we believe that the fundamental forces should act only on the elementary particles. But in particle physics we have been making another assumption as well: that there are mechanisms or principles that pick out which laws are actually expressed in nature, and that these mechanisms or principles also work only at enormously tiny scales, much smaller than the atomic nucleus; an example of such a mechanism is something called "spontaneous symmetry breaking." Given that the choice of laws makes a great difference for the universe as a whole, it began to seem strange to me that the mechanisms that choose the laws should not somehow be influenced by the overall history or structure of the universe at very large scales. But, for me, the real blow to the idea that the choice of which laws govern nature is determined only by mechanisms acting at the smallest scales came from the dramatic failure of string theory.
Like many of the young people trained in elementary-particle physics in the 1970s and '80s, I had great hopes for string theory, since it seemed to have the best possible chance of providing a fundamental unified theory. Indeed, I still think there are ideas in string theory that may be right, and its exploration has led to the uncovering of some beautiful and deep mathematics. But as a theory of the elementary particles, it has certainly so far failed, for while it initially seemed that there was only one possible consistent string theory, we now know there are a great many such theories, each apparently as consistent as the others and all leading to different universes. Thus, string theory hasn't solved the problem of how the world chooses to have the particular collection of particles and forces it does. And whatever the theory's future, I've come to doubt that it ever will.
This crisis led me to wonder whether the search for the principles that determine which laws of nature govern our world could succeed, if we continue to look only at mechanisms that act on very small scales. Instead, I began to ask myself whether there might be mechanisms that could in some way couple the properties of the elementary particles to the properties of the universe created by their interactions — perhaps even on astronomical and cosmological scales. By this I mean nothing mystical. Since the universe has a history, and did apparently pass through a stage when it was very small, there might be some mechanism that coupled the properties of things on the largest scales to the properties of things on the smallest scales. Thus, about five years ago I began to wonder whether there might be some way in which the properties of the elementary particles are chosen by the universe itself, during its evolution. Wondering about this made me notice and take seriously what many people had pointed out previously — that the properties of the elementary particles and the conditions of the universe seem very well chosen for the universe to develop structure and life. It does seem that this is true — that if almost any other set of forces and particles had been chosen, the universe would not only not contain life, it would be much less rich in structure and variety of phenomena than our world is.
Many of the people who've noticed this have become advocates of the anthropic principle. This is the idea that the properties of the world have somehow been chosen because of — or at least are explained by — the fact that with this choice intelligent life like us can exist. I'd always resisted this idea, and I still do. The anthropic principle is said to come in two forms, a weak form and a strong form. In its weak form, I think it's just the observation that the world in which we find ourselves is very special. This doesn't explain anything, it only points out the need for an explanation of how the world got to be special — an explanation that must be made in terms of some mechanism acting in its past. The strong form — that the laws of physics are somehow chosen in order that life can exist — is, to me, really more religion than science. Indeed, I'm not surprised to find that several advocates of the strong form of the anthropic principle are writing books and papers connecting their belief in the anthropic principle with Christian theology. This is fine, for religion, but it isn't science. Instead, when I realized that people like Martin Rees and Bernard Carr were right — that the world is very special in ways that seem a priori extremely unlikely — I began to wonder whether there might be some real mechanism, something taking place earlier in the history of the universe, that might explain how the properties of the elementary particles have been selected so that the world has the enormous amount of structure and variety it does.
At this time, I was reading a lot of biology: Richard Dawkins on evolution, Harold Morowitz on self-organization, and James Lovelock and Lynn Margulis on the Gaia idea. And I remember wondering whether, if the earth can be understood as a self- organized system, maybe the same thing was true for larger systems, such as a galaxy or the universe as a whole. This was also summertime, and I was sailing a lot, and I spent a lot of time letting the boat drift and wondering what kind of mechanisms of self-organization might have acted early in the history of the universe to select the properties of the elementary particles and forces in nature. It seemed to me that the only principle powerful enough to explain the high degree of organization of our universe — compared to a universe with the particles and forces chosen randomly — was natural selection itself. The question then became: Could there be any mechanism by which natural selection could work on the scale of the whole universe?
Once I asked the question, an answer appeared very quickly: the properties of the particles and the forces are selected to maximize the number of black holes the universe produces. This idea came right away, because of two ideas I was familiar with from my work on quantum gravity. The first is that inside a black hole, quantum effects remove the singularity that general relativity says is there — and that we know is there from the theorems of Penrose and Hawking — and a new region of the universe begins to expand as if from a big bang, there inside the black hole. I remember Bryce DeWitt, who is one of the great pioneers of quantum gravity, telling me about this idea shortly after I began to work for him, on my first postdoc. The second idea — which comes from John A. Wheeler, another great pioneer of the field — is that at such events the properties of the elementary particles and forces might change randomly. All I then needed to make a mechanism for natural selection was to assume that these changes are small, because reading Dawkins had taught me the importance for natural selection of incremental change by the accumulation of small changes in the gene. Then, with the universes as animals and the properties of the elementary particles as genes, I had a mechanism by which natural selection would act to produce universes with whatever choices of parameters would lead to the most production of black holes, since a black hole is the means by which a universe reproduces — that is, spawns another.
This was in 1989. I still don't know if the idea is right. But what I'm very proud of is that the idea is testable. Most ideas about why the elementary particles have the properties they do which have been proposed in the past few years aren't testable. This is the main reason the field is in such a crisis. But this idea leads to a prediction, which is that if I could change any of the properties of the elementary particles the result should be either to decrease or to leave alone the number of black holes the universe makes. This is because the idea implies that almost every universe, and therefore most likely our own, has parameters that maximize the numbers of black holes it can make.
When this idea first came to me, I didn't take its prospects very seriously, and I imagine neither did most of my colleagues. I also didn't know much astrophysics, and I imagined that it would be an easy matter to test what would happen to the rate of production of black holes if you changed, for example, the mass of one or another sort of elementary particle, or the strength of one of the forces. So to test the idea, I started to learn some astronomy and astrophysics. So far, I haven't found a way to change the properties of the particles and forces to make a universe that makes more black holes, and I have found several changes that decrease their number. I've also brought the question to a number of astrophysicists, who know the field much better than I do. I've been very pleased that these people, some of whom I admire very much, were interested enough to spend the time to examine such an unusual idea. They made some interesting suggestions, and although no one was able to propose a change of parameters that clearly leads to the production of more black holes, several interesting possibilities, which I'm studying now, did emerge from these conversations. Certainly, if the idea's wrong, I'll be grateful if someone proposes a test that would kill it. I believe more in the general idea that there must be mechanisms of self-organization involved in the selection of the parameters of the laws of nature than I do in this particular mechanism, which is only the first one I was able to invent. But it seems that the situation at present is that there's much more testing that needs to be done, and lately I've been spending more time on this. Perhaps what's most amazing to me is that after five years this rather improbable idea is still not dead.
Whether it dies or not, I've learned enough astronomy to discover something that's completely changed my view of cosmology. This is that the idea that there are principles of self-organization acting on astronomical scales seems really to be true. During the last ten years or so, people who study galaxies have discovered evidence that feedback effects and mechanisms of self-organization are indeed happening at the level of the galaxies; they are, in fact, essential for galaxies to form stars. They're also necessary to the existence of spiral galaxies. The idea that a galaxy is a self organized system — more an ecology than a nonliving clump of stars and gas — has become common among astronomers and physicists who study galaxies.
Thus, it seems to me quite likely that the concept of self- organization and complexity will more and more play a role in astronomy and cosmology. I suspect that as astronomers become more familiar with these ideas, and as those who study complexity take time to think seriously about such cosmological puzzles as galaxy structure and formation, a new kind of astrophysical theory will develop, in which the universe will be seen as a network of self-organized systems.
Beyond this, I also think that — whatever the fate of my ideas — this merging of the science of the fundamental and the science of the organized will overturn the usual ways of thinking about the elementary particles, too. Many of the people who work on complexity, such as Murray Gell Mann, Stuart Kauffman, Harold Morowitz, and others, imagine that the world consists of highly organized and complex systems but that the fundamental laws are simply fixed beforehand, by God or by mathematics. I used to believe this, but I no longer do. More and more, what I believe must be true is that there are mechanisms of self-organization extending from the largest scales to the smallest, and that they explain both the properties of the elementary particles and the history and structure of the whole universe.
To put it most simply, I think a successful theory that merges relativity and cosmology with quantum theory must also be a theory of self-organization. In fact, I have an argument for this conclusion, which is based on the idea that, as Bohr taught us, quantum theory doesn't make sense unless there are clocks and observers in the world. Normally, this is no problem, because the clocks and observers are outside the system being studied, so we can just assume their existence. But if we're going to apply quantum theory to the whole universe, then there's no room for observers or clocks outside the system, because there's no "outside."
But only a complex universe — a universe complex enough to give rise to life — can have things like clocks and observers in it. And if the quantum theory of gravity requires these to exist, and if they are to exist inside the universe the theory describes, then perforce that universe must be complex, and the theory must explain why it's complex. This means there must be some relationship between quantum theory and relativity and self- organization, so that it's logically impossible to describe a relativistic, quantum-mechanical world unless mechanisms of self- organization act in that world to produce the complexity the theory needs if it's to be logically consistent.
A similar argument follows from the way space is described in Einstein's theory of general relativity. For if, as is the case, the only meaningful things in this theory are relationships between real things, then it doesn't make sense to talk about space as being made up of different points, or time as being made up of distinct moments, unless the points and the moments can be distinguished by what's happening there. This means that if it's meaningful in general relativity to speak of the world as having three continuous dimensions of space and one of time, it must be true that the view of the universe from each point of space and time is unique. Otherwise they can't be distinguished from one another. But this means the world must be complex enough so that one can tell where one is in the universe just by looking around. And, again, if the theory of general relativity requires this complexity for its consistency, it must somehow produce it, if it's to be a complete theory of the whole universe.
Thus, I believe that the question of why the laws of physics are chosen so that the world is so complex is intimately related to the basic questions about what space and time are, which we struggle with in quantum gravity. Because of this intimate relationship, I think the next years in elementary-particle physics and cosmology will be very exciting. And what's most encouraging to me is that while many of my colleagues are still depressed over string theory, some of the theoretical physicists whose imaginations I most admire — Alexander Polyakov and Holgar Nielsen, for example — are beginning to look for mechanisms by which the universe could tune the properties of its elementary particles.
Perhaps I might say a word about what it's like to work as a theoretical physicist, because it seems so different from the image I had of it when I was just beginning to dream about it. I don't know if other fields are like this, but what it feels like to work on quantum gravity is that we are on a great search, which is partly one's own search and partly one's participation in a great tradition that's also a wonderful community. Science is a very social activity; we're often traveling, and we spend enormous amounts of time talking with other people — both the friends we work with and people from the larger community. Physics is very verbal. Some of us read one another's papers — I do — but the most important channel of communication is certainly talking. In quantum gravity, there's a community of — I don't know — perhaps a few hundred people who are actively working on the problem and who are in constant communication with one another. Indeed, there's only one thing I don't like about the community of my colleagues, which is that there are still so few women. Of course, there are slowly getting to be more women in the field, but this isn't happening as fast as it is in other fields. It's a very interesting question to ask why this is.
There's another side to doing fundamental science which isn't at all social: it's one's personal confrontation with nature. In the end, I'm trying to understand things like the meaning of time because of my need to know who I am, what this world is, what I'm doing here. To do science is, for me, one kind of response to the alienation of being a small creature in an enormous world. Part of being a scientist, for me, is that in the end I know that I alone am responsible for what I believe.
As a scientist, one can believe what one wants and work on what one wants, but one also accepts the idea that in the end the community is the ultimate judge of the usefulness of what one does. This requires an ethics that makes honesty and respect for the views of others essential, while at the same time making individuality and difference and disagreement essential. So at any one time in the scientific community, there's a consensus about certain matters on which almost everyone has come, after long struggle, to agree; but there's also a large area where no consensus exists. Indeed, this state of affairs is necessary, because if there was too great a consensus the process would stop; this would be the death of science.
Back to Contents 
Excerpted from The Third Culture: Beyond the Scientific Revolution  by John Brockman (Simon & Schuster, 1995) . Copyright © 1995 by John Brockman. All rights reserved.