I was intrigued when Carl Woese told me his collaboration with University of Illinois physicist Nigel Goldenfeld was the most productive one of his entire career, and was pleased to finally run into Goldenfeld last September at lunch in the courtyard of the Santa Fe Institute. But there was little time to talk since I had a meeting in town and he was about to give an "Emergence" lecture.
Goldenfeld, British by birth, speaks faster and more distinctly than most humans, and moves at equally fascinating speed. I stayed long enough at Goldenfeld's lively talk to see him demonstrate various ways carbon organizes, watching first as he pushed over a wooden chair and then got a standing-room-only gathering of some of the world's edgiest scientists to happily agree to put their fingers in the air in an attempt to hold up the ceiling.
More recently while traveling in the Midwest, I planned a visit to the NASA Institute for Universal Biology at the University of Illinois, Urbana-Champaign, where Goldenfeld is director. But flights through Chicago were canceled that day due to a fire at O'Hare, and I was unable to meet with Goldenfeld and the IUB team: Elbert Branscomb, Bruce Fouke, Zan Luthey-Schulten, Charles Werth et al., who are now two years into a five-year $8 million project figuring out the principles of the origin and evolution of life.
Goldenfeld was returning from Washington, D.C. the day of the airport fire and rerouted through Indiana, taking a limo the rest of the way home. He was gracious about our missed appointment and said he and the IUB team clearly don't have all the answers yet, adding that what makes the task particularly tricky is the fact that most life is microbial and that "humans are not representative of general principles!"
As a condensed matter physicist, Goldenfeld's main interest is pattern formation and the processes involved in pattern formation. He researches everything from snowflakes to geological formations to the stockmarket. He and his team have captured gorgeous images of some of these formations in progress at Yellowstone National Park.
A separate curiosity of Goldenfeld's is why so many "offspring" of scientists have autism.
Einstein, Newton, Tesla, and other science giants are now thought to have had Asperger syndrome, a form of autism. And Darwin, as Scott Stossel points out in his recent book on anxiety, exhibited nine of the thirteen indicators of panic disorder (only four are needed to qualify).
Other research interests include high temperature superconductors, turbulence, critical phenomena, polymers and liquid crystals.
Goldenfeld's B.A., natural sciences, and Ph.D., theoretical physics, are both from Cambridge University. He studied there with condensed matter physicist Sam Edwards, who was knighted for his service to science. Goldenfeld says Edwards deserved a Nobel as well -- as did Carl Woese.
Goldenfeld did his postdoc at the Institute for Theoretical Physics in Santa Barbara. He has been a visiting scientist at Stanford University.
Aside from his role at the Institute for Universal Biology, Nigel Goldenfeld is director of the Biocomplexity Group at UIUC's Institute for Genomic Biology. He has been funded by the National Science Foundation for the last 30 years with additional grants from the Department of Energy. He rejects "overtly military" agency funding.
Goldenfeld is the recipient of the Xerox Award, A. Nordsieck Award, the National Science Foundation Creativity Award, among others.
He is a Fellow of the American Physical Society and the American Academy of Arts and Sciences, was an Alfred P. Sloan Foundation Fellow (1987 - 1991), and is a member of the National Academy of Sciences.
Some of the journal editorial boards he serves on include: The Philosophical Transactions of the Royal Society, Physical Biology, and the International Journal of Theoretical and Applied Finance.
He is the author of: Lectures on Phase Transitions and the Renormalization Group, and co-author with Paul Goldbart and David Sherrington of Stealing the Gold: A Celebration of the Pioneering Physics of Sam Edwards.
My phone interview with Nigel Goldenfeld follows.
Suzan Mazur: Carl Woese told me that his collaboration with you was the most productive of his scientific career. What was special about the collaboration for you?
Nigel Goldenfeld: Carl was very generous in saying that. It certainly was a very special collaboration for me. It isn't often that one gets to work with someone with Carl's breadth and intellect.
Carl was, like me, a physicist by training. [Carl Woese, Ph.D. Yale, 1953, Biophysics] He wasn't a theoretical physicist but he had a strong intuition, so we saw things in a similar way. We very quickly found common ground. There didn't need to be a lot of explanation between us. That was quite special because one of the potential obstacles to interdisciplinary collaboration is a language barrier between say biologists and physicists.
Others complain about this but I've actually not run into that too much. People I've worked with in the biological sciences have all been very available to me, and prepared to explain what they're doing in simple language that a physicist can understand. Carl was really open to concepts that I think were important to me as a fairly naive scientist coming in. He was already thinking about the questions that I found natural to ask. They were quite conceptual questions, but concrete ones, and could be answered scientifically. It was clear Carl had been thinking about these questions for decades.
Carl's papers from the 1970s are still fresh because many of the ideas in them have essentially not been picked up by other people. These ideas may have been overlooked because they were not comprehensible to people at the time. They seem very relevant now.
Suzan Mazur: Carl Woese in that same interview with me about two months before he died lamented that he had not managed to overthrow "the hegemony of the culture of Darwin." His view was that a "future biology" couldn't be built within the past structure, that it had to be replaced. And you've both noted that it is "rather regrettable" that Darwin's name is so synonymous with evolution because "there are other modalities that must be entertained."
What are some of these other modalities?
Nigel Goldenfeld: The prejudice that one runs into often (but not always) when one talks about evolution is that evolution really is dominated by what's called vertical gene transfer, where first of all evolution is essentially all about genes -- operations and effects -- evolution occurs because mutations in genes are transmitted to offspring and there's a standard natural selection approach.
There isn't any doubt about this role of population dynamics in biology, by which I mean that entities can replicate and perhaps dominate a population. But the ways in which genetic novelty can emerge and be transmitted are much broader than that, and provide additional channels through which the well-understood mechanisms of population biology can act.
For example, Lamarckian evolution is definitely something that happens in the biological world. As a matter of fact, Darwin himself did not exclude that possibility in Origins of Species. But most people, most lay people and most biologists would regard the idea of something Lamarckian happening as something not within the standard picture of evolution. Lamarckian in this sense really means the inheritance of acquired characteristics.
Suzan Mazur: Epigenetics.
Nigel Goldenfeld: Yes. But there is an even simpler, less controversial example -- horizontal gene transfer.
If you are a bacterium and you receive a plasmid and some genes and you can then later express those genes and transmit them to the daughter cells, that's an absolutely manifest example of something that was acquired in the lifetime of an organism, the result of its environmental interactions that gets transmitted to its offspring. We know that horizontal gene transfer is a significant force in microbial evolution.
Suzan Mazur: So it's happening inside of us, and outside of us as well, horizontal gene transfer because of our microbiota.
Nigel Goldenfeld: Yes. Exactly. That's exactly right. It's responsible in part for the proliferation of antibiotic resistance genes. Antibiotics are becoming less effective as a result of horizontal gene transfer.
Suzan Mazur: So two of the other modalities are horizontal gene transfer and Lamarckism, epigenetics.
Nigel Goldenfeld: Exactly. There's epigenetics and there are other questions. For example, how do we really know that the standard dogma is true that there's genetic variation which occurs as a purely random process and then whatever is present in that variation is then selected by the environment? That would be the standard picture that the evolutionary dynamics itself is random and possible for selection.
Perhaps the best evidence for that is actually a justly-famous experiment: the Luria-Delbruck experiment. It's a brilliant experiment but because of the way the experiment is done, it does not close the book on the idea that organisms and cells can sense environmental stress, and indeed may be able to respond in ways that are not random.
I should add a clarification about what is meant by random, because it frequently gets confused in this sort of discussion. It is generally accepted that mutation rates can be up-regulated due to stress. It is also generally accepted that mutation rates can be site-dependent. Neither of these phenomena relate to randomness in this context: the issue is whether the rate is affected by the environment. As a physicist, for me an experiment that gets a zero or non-zero result is not adequate. If the result is zero (which would be the standard assumption), then one would actually want to bound the result systematically. If not, one would like to measure the dependence as a function of environmental stress. I don't think these sorts of tests have yet been done, and that is why I think there is room to explore this question. Precise tests are beginning to be possible with physics-based tools that can probe living cells in real time.
In short, nobody is saying (well, I'm not at least) that evolution is a teleological process leaning toward a particular end but there's certainly logically a possibility that the evolutionary process is responsive to the environment in a way that is not random. That would not be part of standard Darwinian evolution and population genetics.
Suzan Mazur: Do you differ from Carl Woese on replacing the Modern Synthesis? Woese wrote the following in "A New Biology for a New Century":
"A future biology cannot be built within the conceptual superstructure of the past. The old superstructure has to be replaced by a new one before the holistic problems of biology can emerge as biology's new mainstream."
Nigel Goldenfeld: Our collaborative position was that the Modern Synthesis is simply not enough, population genetics is not a full account of the evolution process because it manifestly does not describe evolution before genes, it does not describe evolution before there were species and the lineages. The Modern Synthesis wasn't designed to do so. Amusingly, Huxley, one of the founders of the Modern Synthesis explicitly wrote that it would not apply to bacteria, but this statement was written before it became clear that the mechanisms of heredity operate in bacteria as in other "higher" organisms.
Usually the way science proceeds is not by demolishing and putting in its place something that's not organically related to it. There's usually an expansion of the theory and a retaining of what still applies, in just the same way that Einstein's theory of relativity supplants Newtonian gravity. It doesn't mean that Newtonian gravity no longer is relevant in the situations for which it is applicable. But Einstein's theory has a greater generality and can apply to physical situations where Newtonian gravity breaks down. The same thing is true with quantum mechanics and classical mechanics.
So our perspective is that if one is looking at origin of life issues, then population genetics, which is really designed for looking at modern life, as it were, is not a full account of what happens generally in evolution and specifically at the early stages of life where genes and species and individual organisms, etc., had yet to emerge. But I don't think there's a real contradiction there, and I don't think this idea is particularly surprising or controversial.
Let me be more precise about the way in which new theory emerges. You might say we have to replace Newtonian gravity by a more general theory, and it has been replaced by Einstein's general theory of relativity. But Newtonian gravity is a perfectly good example of the limit of general relatively in certain circumstances when the gravitational fields are weak and so forth. So I think that's what Carl meant and that's certainly what we were working to try and understand. The Modern Synthesis doesn't address and doesn't claim to address issues of how do living systems even arise in the first place and how do you account for the very existence of life as a phenomenon.
The Modern Synthesis says if you have genes and if you have selection, etc., it gives you a mathematical account of the frequencies of alleles in the populations, and in many situations one could do a calculation. To the extent that can be tested, it does a good job and that's what it was intended to do. That's not the same thing as a comprehensive theory of life as a phenomenon, of the emergence of life, the time before the distinction between genotype and phenotype arose.
So it's replacement with a deeper level of understanding, but it still allows for the old theory in the situations where the old theory applies.
The same thing is true of quantum mechanics. Quantum mechanics was a complete departure from the world of classical physics, yet when we look at systems on a large enough scale and with appropriate energies, the predictions of quantum mechanics are identical to the predictions of classical mechanics, and we can derive classical mechanics out of some limit of quantum mechanics.
That more general theory has a different perspective and a different framework. It has that extra level of generality that we can apply to other situations. I think the same will be true with our understanding of living systems.
Suzan Mazur: Carl Woese also told me that the Last Universal Common Ancestor was not anything material and that evolution is not a "procession of forms," it's more a procession of processes and that physicists and mathematicans are important to the understanding of this new biology. Can you please clarify this idea that LUCA was not something material?
Nigel Goldenfeld: What Carl meant was this. The nature of the evolutionary process we think was different than it is today. Early life was much more collective, much more communal than it is today, particularly the core cellular machinery such as translational machinery, etc., which was horizontally transferred.
We don't know how that happened. It may well have been that there was massive endosymbiosis, meaning organisms were very porous and could crash into each other and absorb each other on a massive scale and that's how cellular functions were transmitted.
That was an idea Carl proposed in the 1970s in one of his papers. But when you have a system which operates in that way, the dynamics -- how the thing changes in time -- obeys very different mathematical principles than what happens post-LUCA when you have vertically-dominated evolution.
An example of this is, suppose you ask what is the defining characteristic of evolving systems, whether they're biologically evolving with genes, etc., or they are other systems you might want to consider living in some sense (for example, financial markets perhaps). The perspective that Carl and I put forward in our last paper together is that life is inherently self-referential. In other words, it's like a computer program which can constantly overwrite the program itself as it runs. The notion that program is the data and the data is the program, the idea that life has a dynamics where the rules that govern life are themselves changed by the rules -- that's self-referentiality. It gives you a kind of description of the physical system that behaves in a mathematical way but is quite unlike any other system that we've ever studied.
It's not like the laws of classical mechanics or the laws of quantum mechanics or the laws of statistical mechanics. Trying to understand the evolutionary process in a more general way requires us to think in a different way mathematically than the way we've thought in the past. I do work with and talk to mathematicians about these sorts of issues.
Suzan Mazur: You're particularly interested in patterns of organization. From snowflakes to the stockmarket to research on autism. Can you tell me a little bit about your ideas regarding each of these areas of research and how they're linked?
Nigel Goldenfeld: I don't know that they are all linked. They're all linked because I'm interested in all of them. It's not that I think there's an overarching principle that connects to say patterns in nature to patterns in ecology to patterns in the evolution of life to autism. That being said, there are some commonalities.
One thing is that systems that are out of equilibrium are much more interesting than systems that are in equilibrium. Systems in equilibrium lapse into perfect states like crystals, etc. Systems that are out of equilibrium are messy and produce turbulence and swirly clouds and human beings and galaxies and strange patterns in space and time. It's interesting to try and understand this, it's my main intellectual interest.
Suzan Mazur: How does autism fit in?
Nigel Goldenfeld: It doesn't really fit in at all. I was interested in autism for many reasons. One is that the autism spectrum conditions are genetic and are more common in offspring of physicists and other scientists and mathematicians than the general population. It's an interesting and important phenomenon.
Suzan Mazur: Asperger syndrome.
Nigel Goldenfeld: Right. When I was on sabbatical at Cambridge years ago I ran into Simon Baron Cohen, director of the autism research center, and I was able to contribute mainly in a technical way to the research they were doing by suggesting new ways to anaylze their data. It's not that there's a grand plan there. But it's sort of connected to my other interests.
Suzan Mazur: Antonio Lima-de-Faria in his 1988 book Evolution without Selection gives a lot of space to crystal organization and growth and notes that the patterns displayed by plants and animals are already present in minerals. What are your thoughts about that in light of your interest in the relationship between geochemistry and metabolism at the scale of atoms in the origin and evolution of life?
Nigel Goldenfeld: I don't remember the parts of that book you're talking about in detail, but it's certainly true that there are very common patterns and motifs that you see over and over again in physical systems and living systems. It's not surprising, because in both those cases what's happening is that you have some variable that may be diffusing, say like heat in a growing crystal or maybe some chemical in the case of something biological or maybe cells diffusing in an environment of food. Those processes have a mathematical aspect but there's not a deeper principle at work.
Suzan Mazur: Lima-de-Faria also refers to bacteria having magnetite particles in their cells and can orient in a magnetic field because of this. Can you say anything more?
Nigel Goldenfeld: I've heard of this but I don't know anything about it. Ralph Wolfe, a wonderful microbiologist who worked with Carl Woese in the discovery of the Archaea, I know he's interested in this phenomenon. It's not true of all bacteria as far as I know.
Suzan Mazur: Do we need a new language for the world around us? Noam Chomsky has said society encourages us to identify what we're looking at as trees, rivers, dogs, etc., when what we're really seeing is something else. So do we need a new language?
Nigel Goldenfeld: Sometimes there are things that are hard to understand because we don't have good concepts to present them and so we have to struggle to find the right mathematical structures to represent them. I don't think we need a new language or know what form that would take. I think the scientific methods will either produce the right language or generate the situations where the need is so acute that it will appear automatically.
Suzan Mazur: Do you see a difference between life and non-life?
Nigel Goldenfeld: Yes.
Suzan Mazur: I've interviewed astrobiologist Maggie Turnbull, for instance, who said the Universe doesn't recognize the distinction.
Nigel Goldenfeld: I think there's a very big difference. The difference is this. When we talk about life, we talk about an abstraction that's occurred at an emergent level of organization. When we look at the individual atoms and molecules that make up life, there's no difference between living and non-living things. On the emergent level, they have different behavior, and the reason they have different behavior is because they have this ability to have a self-referential dynamics. They behave in ways that are very difficult to predict for that reason. So living systems represent a different class of dynamical systems. But at the smallest possible levels, there is absolutely no difference.
Suzan Mazur: Are the factors too overwhelmingly numerous to really come up with a reasonable approach to origin of life? Is it all too relational-based, too unpredicatable, to actually build a self-reproducing protocell? Is origin of life beyond science?
Nigel Goldenfeld: No. I think there are two questions there. One, why does life as a phenomenon even exist? What I mean by that is the following. If you think of anything else that we think we understand -- my pet example is superconductivity because I'm a condensed matter physicist -- there are two levels of understanding.
The first one is why the phenomenon exists and is it based on very general principles of symmetry and conservation, conservation of energy, gauge invariance, etc. For superconductivity, that level of understanding came very late, and its absence prevented us from discovering many of the core features of superconductivity for several decades. I'm thinking of the expulsion of magnetic fields, for example, or the existence of what is known as type-II superconductivity and the existence of magnetic vortex lines. Once we understand why the thing can exist in the first place, then we can understand how it is instantiated in any particular system.
For example, for classic superconductors, we understand the connections between the electrons and the sound waves in that particular material and how it is all put together quantum mechanically with electricity and magnetism. So for these materials we understand superconductivity both at a general level from the point of view of symmetry and symmetry breaking, and we also understand it at a level of specific behavior of the atoms and molecules in the crystal structure. For the high temperature superconductors, we still lack this level of understanding of instantiation, although we know they follow the same basic principles as the classic superconductors.
When we talk about living systems we don't have the general understanding, the symmetry-based understanding. We're unable to say, yes, this is a phenomenon and should exist. We need to develop a fundamental understanding of living systems, and from that we can go on to ask, well, given that life exists, how does one instantiate it? How does one make examples in the lab as it were? Up to now, most biological research has been akin to studying superconductors from the materials perspective only, lacking the more general symmetry-based level of understanding. I believe that just as the absence of a general theory held back the development of superconductivity as a quantitative science for decades, the absence of a general theory of life is holding back our understanding of biology. Where this manifests itself mainly is in how we intervene with biological systems, medically and ecologically. The emergence of drug resistance in cancer, or herbicide resistance for that matter, are both examples of our chronic inability to control these systems.
You asked if the origin of life is beyond science. Jack Szostak and others are trying to make synthetic cells, etc. I think science can constrain the environment where life on Earth first arose and in principle be predictive about what sort of life might exist elsewhere in the Universe and what might be the properties and biosyntheses. It's not beyond science. It's a question that science can definitely answer.