Complexity theory guru or great British eccentric? Lucy Hodges meets Stuart Kauffman who simulates the origins of life on a computer

As a Marshall scholar at Oxford in the early 1960s, Stuart Kauffman learned the art of inventing things on the spot.

By sitting at the feet of the eminent philosopher Geoffrey Warnock and psychologist Stuart Sutherland, now emeritus professor of psychology at Sussex University, he received an education that has stood him in good stead. In his book At Home in the Universe, to be published later this month, he explains how his Oxford professors asked questions such as "Was language prior to recognition?" and "How could a neural circuit allow the eye to distinguish the offset of two parallel lines when the distance between them was less than the width of a single cone or rod in the retina?" One got training at invention, he writes, a training nurtured by an English tradition worth emulating: the British love their eccentrics.

Today Kauffman, aged 55, is following in this tradition as one of the foremost gurus at the Santa Fe Institute. He is one of the most ambitious of the new complexity theorists, contenting himself with nothing less than the origins of life.

ÖйúAƬ

Trained as a medical doctor, with a first degree in philosophy, Kauffman has spent much of his career trying to rewrite Darwin's theory of evolution, the notion that living things have evolved by the fittest of each generation surviving to pass their characteristics on to their offspring. Natural selection, the process whereby the environment randomly "selects" certain organisms, whose genes are thus more prevalent in the next generation's gene pool, cannot on its own explain the origin or the subsequent evolution of life, says Kauffman.

Using elaborate computer simulations, he suggests that the way the world is ordered is not accidental. It is not entirely the way it is through random variation. "Laws of complexity spontaneously generate much of the order of the natural world," he says in his new book. "It is only then that selection comes into play, further moulding and refining."

ÖйúAƬ

We know that simple physical systems show spontaneous order, he explains. An oil droplet in water forms a sphere; snowflakes exhibit an evanescent sixfold symmetry. "What is new is that the range of spontaneous order is enormously greater than we have supposed. Profound order is being discovered in large-complex, and apparently random systems. I believe that this emergent order underlies not only the origin of life itself, but much of the order seen in organisms today".

The actual science is done by theoretical modelling. At the Santa Fe Institute Kauffman has created chemical "networks" on computer screens. In these networks molecules (polymeric macromolecules) react together. They are controlled by catalysts and generate products that may in turn create further reactions. Sufficiently complex mixtures of such polymers can replicate as a group, says Kauffman, even if no single member of that group can replicate by itself. He thinks that, if you start out with a "complicated enough" soup of chemicals, the chemistry needed to reproduce life is bound to emerge. "Life is a natural property of complex chemical systems," he says. "When the number of different kinds of molecules in a chemical soup passes a certain threshold, a self-sustaining network of reactions an autocatalytic metabolism will suddenly appear."

All of which is extremely controversial. Peter Coveney and Roger Highfield, in their new book, Frontiers of Complexity, quote Leslie Orgel, of the Salk Institute, San Diego, suggesting that chemicals do not behave in the way that Kauffman describes. Asked about such criticism, Kauffman says it is not quite fair. "I think that one is going to be able to get collectively autocatalytic sets of molecules and the molecules might be all kinds of polymers," he replies. "Nobody has made a collectively autocatalytic set yet, but cells are collectively autocatalytic sets and they're made up of DNA and RNA and proteins and all sorts of small molecules and so on.

"It's not just the template-replicating properties of DNA that makes a cell as a whole reproduce. DNA can't do it by itself. It needs RNA and proteins and everybody knows that. So free-living organisms are collectively autocatalytic systems with many more kinds of molecules than DNA alone or RNA alone, and while many people believe that the template-replicating properties of DNA are essential to achieve self-reproduction, I don't." Laughing, he adds: "I respectfully disagree with Leslie. He's a superb chemist, and I'm certainly not, but I just respectfully disagree."

Kauffman may not be a superb chemist but he is a pretty impressive polymath. After his philosophy degree at Dartmouth, he studied philosophy, psychology and physiology at Oxford, decided he did not want to be a philospher, and opted for medicine. On his way to medical school he stopped for a year at Berkeley where he took a course in embryology and fell in love with developmental biology. He became interested in the problem of cell differentiation. It had just been discovered that genes could turn one another on and off like neurons do. Effectively, genes were part of a circuit, so it was possible to have different steady-state patterns of gene activity.

He pondered the following question: what would it take to get orderly behaviour in a genetic system with, say 100,000 genes, which is roughly the number of genes in a human cell? How hard would evolution have to work? Would it have to be a very precisely crafted thing like a Swiss watch or some finely crafted computer programme that was wrought by natural selection? "Or might it be the case that if you just threw everything in together haphazardly, more or less randomly, that spontaneous order would emerge?" he says. "And, if so, Mr Darwin's motor of natural selection wouldn't have to work quite so hard."

ÖйúAƬ

Kauffman set out in medical school to see if that might be true. Which shows how long he has been at this particular task. In fact, he started out on this road at Berkeley 31 years ago on his pre-med course when he began to invent genetic circuits in little notebooks. "I remember the passion with which I was wondering 'Could you get order spontaneously or would it have to be the case that it was all carefully crafted by natural selection'," he says now.

"And I just wanted it to be true that one could find order arising even in randomly built systems. And you can imagine my delight when it turned out that that was right. I was just stunned." At medical school at the University of California at San Francisco, he made random networks of lightbulbs turning one another on and off as models for genetic networks on a computer, and discovered that there was in fact a spontaneous order.

ÖйúAƬ

Perhaps unsurprisingly, Kauffman wound up winning the research prize in his class at medical school. From there he moved effortlessly as assistant professor to the University of Chicago in the department of theoretical biology. There he began his theoretical and experimental work on what he calls "the origin of life model", based on autocatalytic sets. But at that time he also became diverted by fruit flies, and ended up spending ten to 12 years working on research into these organisms.

There followed a spell at the National Institutes of Health, specifically the National Cancer Institute, where he was fulfilling military requirements, this being the Vietnam War era. Once that was over, he took up a position at the University of Pennsylvania in the medical school, where he spent the next 19 years. Much of his work during those years was on the cell cycle he was busy trying to show that the cell cycle is a chemical oscillation and on how the patterns of early embryo genesis are set up in the developing egg. In the early 1980s he attended a lecture in Boston on evolving novel genes in bacteria to catalyse new reactions. As a result he had a brain wave. You could clone random DNA sequences into bacteria make random genes and random proteins by the billions and find out how hard it is to make random protein catalyser reactions. This was the now burgeoning area of molecular diversity with which Kauffman became fascinated.

How come he ended up in living in what many regard as one of the world's most sublime spots, Santa Fe, New Mexico? It was a combination of factors, he explains, some engineered, others accidental. First, he fell for Santa Fe in a big way. Second, the Santa Fe Institute was just setting up. And, third, he and his wife bought a house there. The trouble was he was still living and teaching in Philadelphia. Tragedy struck. His beloved 13-year-old daughter was killed by a hit-and-run driver. Kauffman was devastated. He began to spend half his time in Santa Fe, which helped. But what helped further was being awarded a Macarthur fellowship, otherwise known as a "genius" award. These are a big deal in America. They are awarded to very few, highly original people, and they are generous, intended to free up the creative spirit for a five-year period. "That was very nice,'' he says, modestly, of being given an award. The grant enabled him to stay in Santa Fe. "It's a sort of very Macarthur fellows sort of thing to hang out with all these people in this neat think tank in Santa Fe," he explains.

His semi-popular book now hitting the bookshops, written in a non-scientific style, and with many personal references, is an attempt to reach a non- scientific audience. In America it has been taken as a Book of the Month Club alternate choice for Christmas, but what surprises the author is that the business community is so interested in it. The management consultants, McKinsey, reproduced a chunk of the book in its quarterly magazine and the accounting firm, Ernst and Young, has bought 1,500 copies to distribute to its staff. That is because Kauffman has things to say about the economy as a web of technologies, and about fitness landscapes and learning curves. The theory of fitness landscapes is that as you climb higher and higher it is harder to find ways to get uphill. Similarly with learning curves: at first you improve rapidly, and then ever more slowly. "It's catching the attention of a fair number of business folks," he says. "It may turn out to be really useful."

Like other complexologists Kauffman is hoping that the science of complexity will reveal new laws to us. We have had 2000 years of reductionist science and the time has come to put the pieces together, he says. The modern tool for such study is the computer which is a kind of macroscope, doing for today's generation what the telescope and microscope did for yesterday's, but in reverse. "The telescope and microscope revealed new worlds to us," he says. "The computer as a macroscope is allowing us to see what sort of emerging collective phenomena fall out from complicated systems made up of relatively simple parts, but lots of them.

"I have a hunch that we really are entering an exciting era in which we are going to be discovering all kinds of laws about collective emergent phenomena' It's really true that we will see ourselves anew in the universe, he thinks. All of which sounds plausible enough, if one accepts that complexity theory is more than highly sophisticated computer hacking. Towards the end of our conversation, Kauffman becomes even more ambitious. There is a spiritual theme running through his new book. In our search to understand the world, we have lost our paradise, he says. "Paradise has been lost, not to sin, but to science." Yet we hunger for things spiritual. Kauffman is hoping that through the new science of complexity we may recover our sense of worth and our sense of the sacred.

ÖйúAƬ

At Home in the Universe: The Search for Laws of Complexity, by Stuart Kauffman, published this month by Penguin UK.