Chapter 19 life and death

life and death After Arthur arrived for lunch on the first day at the Santa Fe Institute, he and Kaufman walked along the adobe gallery on Canyon Road to Kaufman's favorite water hole.In the two weeks since then, they have had lunch together, or just talked, almost every day. They often chat while walking.Kaufman liked the fresh air outside even more than Arthur.He participated in numerous excursions and backcountry camps in Sierra as a Boy Scout boy.In college he was a vigorous skier and mountaineer.Now he still likes to go hiking whenever he can.So Kaufman and Arthur were always walking and talking along the Canyon Road, or climbing up the open hill behind the Abbey.They sat on top of a hill overlooking the panorama of Santa Fe and the rolling mountains.

Arthur began to feel what seemed to be an unspeakable sadness in Kaufman's heart.Sometimes, in the midst of his witty, witty, inquisitive conversations, as he eloquently discussed his thoughts, he would stop abruptly and a look of sadness would flicker across his face.One evening not long after Arthur and his wife, Susan, arrived in Santa Fe, the Arthurs and the Kaufmans went out to dinner.Kaufman told them a sad story: He and Lisa came home one Saturday night in October of that year to learn that their thirteen-year-old daughter, Merritt, had been hit by a car and fled in a car. No, their daughter was taken to a local hospital with serious injuries.The couple and their son Eisen ran to the hospital immediately, but when they arrived at the hospital, they were told that Merritt had passed away fifteen minutes ago.

Today, more than five years after the incident, Kaufman can't help himself when he recounts the incident.But that night he couldn't help himself.Merritt is his most beloved daughter. "This disaster has literally broken me. I am devastated beyond words. We went upstairs and my daughter's broken body was lying on a table, cooling. The grief was unbearable. The three of us huddled together in bed that night, weeping. My daughter has an excitable personality, but we were amazed by her caring nature. We all thought she was the best of the four of us." Kaufman continued: "People say that time heals wounds. But that's not quite the case. It's just that grief flares up less often."

As they walked along the road and climbed the hillsides surrounding the abbey, Arthur couldn't help being fascinated by Kaufmann's concepts of order and self-organization.But the irony is that when Kaufman uses the word "order," it is clearly the same thing that Arthur used "chaos"—that is, emergence, that is, the perpetual process by which complex systems organize themselves. trends in various forms.But at the time, it was not surprising that Kaufman used the exact opposite.He happened to come from a completely different direction with this concept.Arthur talks about "chaos" because he starts with the frozen abstract world of economic equilibrium, where the laws of the market are thought to determine everything with as much precision as the laws of physics.Kaufman talks about "order" because he starts with the messy and accidental world of Darwin.In Darwin's world there are no laws, only accidents and natural selection.Although they started from completely different angles, they basically arrived at the same position.

Kaufman was also fascinated and puzzled by Arthur's concept of increasing returns. "It's hard for me to understand why this concept is new in economics, when biologists have been dealing with positive feedback for so many years." It took him a long time to figure out how stagnant and rigid the views of the world of neoclassical economics were . But what got him even more interested was that Arthur began asking him another economics question that was fascinating him: technological change.To put it lightly, this issue has long since become a hot political issue.You can sense this underlying anxiety in any newspaper or magazine you flip through: Is America still competitive?How did we lose the mythical American creativity, the skills of the old Yankees?Will the Japanese squeeze us out one industry at a time?

These are all straight-to-the-point questions.The point is, as Arthur explained to Kaufman, economists cannot answer these questions, at least in the most basic theory.The whole dynamic of technological development is like a black box. "Until fifteen or twenty years ago, people still believed in the consciousness that technology just randomly fell from the sky, and humans emerged from the blueprint of the blueprint to produce steel, manufacture silicon wafers, or produce any other such thing. These technologies were invented by smart people like Thomas Edison. These smart people were inspired by the sky while lying in the bathtub, so they can add a page to the blueprint drawn by the sky.” In fact, strictly speaking, technology It's not economics at all.Technology is foreign, magically conceived by non-economic processes.Recently, a lot of research has been done using simulations to demonstrate that technology is intrinsically conceived, which means that technology is generated from within the economic system.But that often means technology is the result of investment in research and development, almost like a commodity.Although Arthur believed that this view contained a certain truth, he still did not think it was the essence of the matter.

He told Kaufman that when you look at economic history that runs counter to economic theory, technology is less like a commodity and more like an evolving ecosystem. "Technological inventions, in particular, are rarely created in a vacuum. For example, a laser printer is basically an electrostatic printer, which is a laser device and computer lines to tell the electrostatic cylinder where to print. So, only when we have computer technology , laser technology and xerography technology, the laser printer will be invented. But the laser printer will not be invented until people need compact, high-speed printers."

In short, technology forms a highly interconnected web, or in Kaufman's term, a network.What's more, these technological webs are highly dynamic and unstable.Technology seems to evolve like living things, just as laser printers produced desktop publishing software, and desktop publishing software opened up a world for graphics processing programs.Arthur said: "Technology A, B, and C may lead to the possibility of technology D, and so on. In this way, a network of possible technologies is formed. In this network, various technologies fully penetrate each other, develop together, and generate more and more technological possibilities. And just like that, the economy becomes more and more complex.”

Moreover, these technological webs, like biological ecosystems, experience bursts of evolutionary creation and mass extinctions.For example, new technologies such as automobiles have replaced the old technology of using horses for transportation, and with the disappearance of the way of using horses for transportation, blacksmith shops, fast carriages, water tanks, stables, horse breeders, etc. have also disappeared. The entire technological subsystem of the method suddenly collapsed in what the economist Josef Schumpeter once called a "hurricane of destruction."With the automobile came paved roads, gas stations, fast food restaurants, motels, traffic courts, traffic police, and traffic lights.New goods and services began to proliferate, each new insertion because space was made for them by goods and services that had appeared before.

Indeed, says Arthur, this process is an excellent example of what he refers to as the rate of increasing returns: Whenever a new technology opens up suitable space for other goods and services, the people who enter that new space There is a great temptation to do what you can to help this technology grow and thrive.What is more, this process is the main driving force behind the phenomenon of lock-in: the more new space a given technology can provide to other technologies attached to it, the harder it is to change the direction of this technology development, unless there is a Much more powerful technology emerges.

Arthur explained that the concept of this web of technology is very close to his vision of the new economics.The problem was that the mathematical methods he developed were only suitable for observing the development of one technology at a time.What he really needed was a network-like simulation like the one Kaufman had developed.So he asked Kaufman: "Can you do a simulation where a technology that's just been invented acts like a switch that's turned on, maybe...?" Kaufman listened dumbfounded to hear all this.can heWhat Arthur had just described in a completely different language was exactly what Kaufman had been working on for the past fifteen years. After a few minutes of meditation, Kaufman began to explain to Arthur why the process of technological change is like the origin of life. Kaufman first came up with the idea in 1969, when he arrived at the Theoretical Biology Research Group in Chicago. After medical school, being in Chicago was like being in heaven, he said.In retrospect, Chicago was actually the second of the three most exciting intellectual environments he experienced. "It's an extraordinary place with extraordinary talent," he said. "The department where I work in Chicago has as good a collection of people as any other in the United States, and they are like the group of friends I met in Italy." Jack Cowan is making a breakthrough in cortical organization. research work.He used simple equations to describe how waves of stimulation and inhibition in the brain travel across double-scale sheets of nerve cells.John Maynard Smith is also doing groundbreaking research on evolutionary dynamics.He used a mathematical technique known as game theory to clarify the nature of competition and cooperation among species.Maynard, who came here to do research on the annual leave of the University of Sussex, gave Kaufman timely help in the mathematical analysis of the network. "John taught me arithmetic, that's what he said," Kaufman said. "One day I cured him of pneumonia." Living among colleagues and close friends, Kaufman soon discovered that he was not alone in his study of the statistical properties of the web.For example, in 1952, the English neurophysiologist Ross Ashby considered the same question in his book Design for a Brain."He was exploring the ubiquity of complex networks, asking a question that was similar to mine, but I didn't know anything about it," Kaufman said. "As soon as I found out about it, I immediately got in touch with him." connect." At the same time, Kaufman found himself making some cutting-edge extensions to physics and applied mathematics while studying genetic networks.The dynamics of his gene-regulatory network turned out to be a special case of what physicists call "non-linear dynamics."From a nonlinear point of view, it is easy to see why sparsely connected networks can self-organize stable cycles so easily: they behave mathematically as if all the rainwater falling on the hillsides around the valley would It flows into the lake at the bottom of the valley.In the space of all possible network behaviors, stable loops act like basins, or "attractors," as physicists call them. After six years of painstaking research into genetic networks, Kaufman felt satisfied that he finally understood their mysteries so perfectly.But he still couldn't help feeling that something was missing.The theory of self-organization of gene regulatory networks is of course very good, but at the molecular level, gene activity depends on complex, delicate molecular mountains of ribonucleic acid (RNA) and deoxyribonucleic acid (DNA).And where did RNA and DNA come from? How did life originate? According to the orthodoxy in biology textbooks, the origin of life is fairly straightforward. DNA, RNA, proteins, polysaccharides and other molecules that form life formed billions of years ago in a small warm pond.Back then, simple molecular building blocks like amino acids had to accumulate in the nascent climate.In fact, in 1953, Nobel laureate Harold Urey in chemistry and his graduate student Stanley Miller demonstrated experimentally that, initially, methane, ammonia, and the like The resulting atmosphere can spontaneously generate such construction bricks.All it takes is the occasional lightning bolt to provide the energy to create the chemical reaction.The theory goes on to say that over time these simple mixtures begin to accumulate in ponds and lakes, undergo further chemical reactions, become more and more complex, and eventually produce a population of molecules, including the DNA double helix and (or) its single-struct cousin RNA.They all have the ability to reproduce themselves, and when self-reproduction occurs, everything else is the result of natural selection.Conventional biological theory roughly says this. But Kaufman doesn't buy it.Just one thing first, most biomolecules are very bulky in structure.For example, the synthesis of a single protein molecule requires the precise orderly assembly of hundreds of amino acid building blocks.This is very difficult to achieve in a modern laboratory with all the most advanced biochemical techniques.And how can protein molecules form themselves in a pond?Many people have tried to calculate how random this could be, but they have come to basically the same conclusion: if the formation of biomolecules is really random, you have to wait much longer than the universe formed It is only possible to wait for the formation of a useful protein molecule, let alone how long it will take for the formation of countless proteins, sugars, fats and amino acids to combine into a fully functioning cell.Even if you assume that planets with warm oceans and climates like Earth exist among the trillions of planets in the millions of galaxies in the observable universe, the possibility of life emerging on any of these planets is still is negligible.If the origin of life was really a random result, then the origin of life is really a miracle. More specifically, Kaufman disagrees with the standard biological theory because it equates the origin of life with the emergence of DNA.To Kaufman, basing the origin of life on something so complex is clearly unreasonable. The DNA double helix is ​​capable of self-replication, of course, but the key is that this ability to self-replicate lies in its ability to unfold its own double helix and make copies of itself.What's more, in modern cells, this process also depends on a group of special protein molecules that play various assisting roles.How did all of this happen in one pond?Kaufman said: "I had the same urge to find out if there is order in the gene regulatory network. There is something very wonderful in DNA, and the origin of life depends on this one. Something special, I just don't want it to be true. I said to myself: 'What if God gave nitrogen another valence? (full of nitrogen atoms in the DNA molecule) If so, life would be possible Is it?' It was a shocking conclusion to me that the origin of life should have been on such a delicate balance." But then Kaufman thought, who said that the essential factor of life is DNA?From this point of view, who said the origin of life is a random result?Perhaps there is another way to produce self-replicating systems, one that would allow living systems to evolve from simple chemical reactions to life on their own. Well, now imagine the original stock, with those tiny amino acids, sugars, and so on.Obviously, you can't expect them to fuse together by themselves to form a cell.But you can at least count on some random interactions between them, and in fact, it's hard to imagine anything that would prevent them from interacting like this.While random interactions don't produce anything fancy, they are able to produce larger numbers of small molecules with short chains and forks. Although this fact does not increase the possibility of the origin of life at present, Kaufman thinks, hypothetically, only hypothetically, that there are some small molecules floating in the initial soup that can act as "contact agents", which is like extremely Tiny matchmakers.Chemists can often find substances in which one contact agent molecule sticks to two other molecules as it travels around, bringing them together so that the interaction and fusion between them develops more quickly.The contact agent then lets go of the "newlyweds" and brings the other two together, and so on.Chemists are also very familiar with the many axe-like contacts molecules, which sideways on one molecule after another and then cut them apart.These two roles of contact agents make them the backbone of the modern chemical industry.For example, gasoline, plastics, dyes, pharmaceuticals, etc., without contact agents, all these products are almost impossible to appear. Well, Kaufman thought, now imagine some A molecules in the initial stock soup busy catalyzing the formation of another B molecule.Since the first molecule was formed randomly, its contact and catalytic functions may not be very efficient, but it doesn't have to be very efficient.But even a weak contact agent can make the formation of B molecules much faster than the other way. Now, Kaufman thought, let's assume that molecule B itself is weakly contact-catalyzed, so that it catalyzes the production of some molecule C.Suppose molecule C can also act as contact catalysis, and so on.He speculated that if the pond of primordial soup was large enough, and if the various molecules in the pond were sufficiently numerous to begin to interact, it might be entirely possible, at some stage in the development of events, to produce molecules that had completed the circle, and then Start going to Molecule Z that catalyzes Molecule A.But then there's more molecule A, which means more contact catalyst for enhancing the formation of molecule B, which in turn is available for enhancing the formation of molecule C, and so on and on and on Proceed with it. In other words, Kaufman realized that if the conditions in the initial soup are right, you don't have to wait for random effects at all.The mixture in the initial stock forms a coherent, self-reinforcing web of interactions.Furthermore, each molecule in the net is able to contact and catalyze the formation of other molecules in the net, so that all the molecules inside the net are steadily growing larger and larger than the molecules outside the net.In short, viewed as a whole, this web can catalyze the formation of the self.It will be an "autocatalytic group". There was a sense of awe as Kaufman realized all this.Here again order emerges, the order of free existence.Order arises naturally from the laws of physics and chemistry.Order emerges spontaneously from molecular chaos, announcing itself as a developing system.The idea is unbelievably good. But is this life?No.Kaufman had to admit that this was not life as we know it today.An autocatalytic group has no DNA, no genetic code, and no cell membrane.In fact, it doesn't really exist on its own, except as a group of molecules floating in a primordial pond.If a Darwin outside the Earth's atmosphere happened to pass by at that time, he (or it) would hardly have noticed anything unusual.Any given molecule participating in this autocatalytic group looks like any other molecule.We cannot find the essence of the matter in any one autocatalytic group, the essence of the matter lies in the overall dynamics of this autocatalytic group: its collective behavior. In a deeper sense, however, Kaufman thought, the autocatalyst might be alive, capable of exhibiting some very distinct signs of life.For example, it can develop.And, in principle, there is no reason why such autocatalytic groups could not be open-ended, capable of producing more and more complex molecules over time.Such autocatalytic groups even have a metabolic function: a network of molecules that stabilizes the supply of amino acids and other forms of molecules floating throughout the initial broth as "food" molecules, glues them together, and turns the autocatalytic Catalytic groups become more complex hybrids. Autocatalytic groups can even display primitive ways of self-reproduction: If an autocatalytic group happens to be splashed from a small pond into an adjacent pond, say, during a flood, then the The autocatalytic group immediately begins to develop in the new environment.Of course, if another autocatalytic group already existed in the pond, the two groups would compete for resources.This, Kaufman realized, opened the door directly to natural selection to sublate and optimize these autocatalytic groups.We can easily imagine such a process of natural selection.Autocatalytic groups that are more adaptable to environmental changes, or have more effective contact catalysis, are better at generating interactions, or have more complex and refined molecules have been preserved by natural selection.In the end, in fact, as you can imagine, the sublation process produced DNA and everything else.The key is to first form an entity that can survive and reproduce itself.After that, evolution was able to do its job in a relatively short amount of time. Well, he admits that this is an assumption, many ifs on top of many ifs.But to Kaufman, the story of this autocatalytic group runs counter to the most eloquent explanation of the origin of life he's heard.If his hypothesis is true, it means that the origin of life did not need to wait for some absurd and impossible event to produce an extremely complex set of molecules.This means that life may indeed develop from very simple molecules into existence by its own efforts.This also means that life is not a random accident, but a manifestation of nature's self-organized, continuous compulsion. Kaufman is simply obsessed with this research.He immediately plunged into calculations and computer simulations of random networks, repeating the experiments he had done at Berkeley: he hoped to understand the laws of nature in the autocatalytic group.Well, even if you don't know exactly what kind of mixtures and what kind of chemical reactions there were at the time, you can at least imagine their possibilities.Is the formation of autocatalytic groups a complete improbability?Or was its formation almost inevitable?Let's look at the data.Suppose there are a small number of "food" molecules, say something like an amino acid, and suppose that in the initial stock, these molecules begin to aggregate with each other, forming chains of polymers.How many polymers can be polymerized in this way?How many interactions must occur between these polymers to form a large network of interactions?If such a large network of interactions is formed, how likely is it that an autocatalytic group forms after autism? "When I thought it through, it became obvious to me that the number of interactions would be greater than the number of polymers. In this way, each polymer would be able to catalyze a reaction after reaching this fixed possibility, it reaches some kind of mutual autocatalyzed complexity phase. In other words, it's like his genetic network: if primordial soup crosses the line of complexity, it goes through this comical phase change that Phase transition. Then the emergence of autocatalytic groups is indeed inevitable. In a sufficiently rich initial soup, autocatalytic groups can only be formed in this way, and life spontaneously bonded and catalyzed out of the initial soup.” Kaufman felt that the story was too beautiful to be true.He said: "I still believe in this plot of the story as much as I did when I first came to the conclusion. I believe that this is how life begins." Arthur also agrees with Kaufman's point of view.He thought it was a great discovery, not only because it was a brilliant explanation for the origin of life, but also because the autocatalyst was so similar to economics that he simply couldn't ignore it.He and Kaufman riffed on this point in those days as they walked the hills or hunched over their lunch table. They all agree that, most obviously, the autocatalytic group is a molecular switching web, just as the economy is a switching web of goods and services.In a real sense, the autocatalytic group is actually like a tiny economic system that takes raw materials (primitive "food" molecules) and turns them into useful products (that is, more in the autocatalytic group molecules). What's more, the autocatalytic group can evolve on its own, like an economy that can grow more complex over time.This is what fascinates Kaufman the most.If inventions are new combinations of old technologies, then the number of possible inventions increases dramatically as we have more and more old technologies at our disposal.In fact, he argues, once things pass a certain threshold of complexity, a phase transition similar to the one he discovered in the autocatalytic group occurs.And below the complexity critical point, some countries rely on only a few types of industrial production, and the economies of these countries tend to be fragile and stagnant.In this case, it doesn't matter how much investment you pour into the country. "If you just grow bananas, you've got nowhere else to go but more bananas." But if a country starts trying to diversify, increasing the complexity of the economy beyond critical At that point, the country enters an explosive phase of development and invention—what some economists call the "economic takeoff" phase. The existence of phase transitions also helps explain why trade is so important to economic prosperity, Kaufman told Arthur.Suppose there are two different countries, and each country's development is below the critical point, and the economies of these two countries will not improve.But if these two countries start to trade, their respective economies will enter a stage of interdependence, forming a more complex and larger economic system. "I believe that the trade between these two countries can form a joint economic system, so that the tipping point can be surpassed, and the economy can expand outward like an explosion." Finally, an autocatalytic group is perfectly capable of going through the booms and busts of evolution like an economy.Injecting a new molecule into the initial soup can often completely change the structure of the old autocatalytic group, which is the same reason that the economic system changed when the horse instead of walking was replaced by the appearance of the automobile.This is what really appealed to Arthur about autocatalysis.The same feature in autocatalysis fascinated him as it did when he first read about molecular biology: disturbance and change, and some serious consequences, originate in seemingly trivial things.Behind these phenomena lie profound laws of nature. Kaufman and Arthur had a great time discussing these ideas and exploring connections, endlessly.Their conversation is like a free discussion of first-year college students anytime and anywhere.Kaufman, in particular, was thrilled.He felt they were exploring something really new.Obviously, web analytics won't help anyone predict exactly what new technology will come out next week.But it may help economists gain statistical and structural means of forecasting this process.For example, when you introduce a new product, how much does it shake the economy?How many other goods and services will it lead to?How many old industries will it kill?How do you recognize that a commodity has become the center of an economic system, not just another hula hoop? Kaufman further recognized that the utility of these ideas could ultimately extend far beyond the field of economics. "I think these models can also accommodate accidents and laws. The point is that there may be regularities in phase transitions, but there are no regularities in the specific details. Maybe we have a grasp of how the process of historical development begins. Models that explain the origins of historical events such as the Industrial Revolution or the Renaissance as a cultural shift, why a closed and isolated society or social psyche cannot remain closed and isolated after some new infusion of ideas.” You can also The Cambrian explosion asks the same question: 570 million years ago, the world of algae and pond scum suddenly exploded into a world of complex, multimolecular organisms. "Why the sudden diversity?" Kaufman asked. "Perhaps the explosion was caused by the world passing a tipping point of diversification, or perhaps the world developed more nutritious and complex matter from the kelp mats, which resulted in one transformation process leading to another transformation process Explosive periods. It’s the same economic phenomenon.” Of course, even Kaufman has to admit that all these thoughts are nothing more than hopes.But on the other hand, he told Arthur, it might be very possible.He has been doing basic research since 1982, when he resumed research on autocatalysis after a hiatus of more than ten years. Kaufman remembers that he stopped studying autocatalysis on a certain day in 1972.At that time, a chemist in Chicago, Stuart Rice (Stuart Rice) visited his theoretical biology group.Rice had a great reputation in theoretical chemistry, and Kaufman was eager to impress him. "He came into my office and asked what I was working on, and I told him I was working on autocatalysis, and he was like, 'What are you working on?' I don't know why he said that. I guess he thought it was The work didn't make any sense. But at the time I thought: 'God, of course Stewart knows what he's talking about. I shouldn't be doing this anymore.' So in 1971 I wrote up what I had done and published it in Cybernetics Society journal, then put the study aside and forget about it." Kaufman's reaction wasn't all about insecurity.In fact, his autocatalytic model happened to have reached a dead end at that time.No matter how many calculations and computer simulations he does to study the origin of life, they are just calculations and computer simulations.To make real progress, he would have to build on the experiments of Miller and Urey, and he would have to prove in the laboratory that autocatalytic groups could indeed be produced in the initial broth.But Kaufman didn't know how to do this.Even if he had the patience and the skill to perform chemical experiments in the laboratory, he would have had to observe the formation of millions of mixtures in various combinations at various temperatures and pressures.It would be something he would spend his entire life with no results. No one seems to be able to figure out what to do.Kaufman is not alone in autocatalysis research.Several years earlier, in his 1969 book Chemical Evolution, Berkeley Nobel laureate Melvin Calvin described several of the different autocatalysts he had detected for the origin of life. situation.At the same time, Otto Roseeler and Manfred Eigen in Germany were also independently exploring autocatalysis.Eigen has even been able to demonstrate a form of autocatalytic cycle in the lab using RNA molecules.But no one has yet been able to demonstrate how the autocatalytic group emerges from the simple molecules in Miller-Urey's initial soup.There doesn't appear to be any progress on this pending doctrine. But while Kaufman's reaction to Rice's comments wasn't entirely insecure, it was largely.He had an urgent need to prove his abilities in this new field, but he found that theorists had a low reputation among biologists. "In biology, the people who do the math are at the bottom of the bottom," he said.This is the exact opposite of what happens in physics and economics.In physics and economics, the theorist is king.In the field of biology, especially in the fields of molecular biology and developmental biology, experimental tools are brand new, and a large amount of data needs to be collected in order to study the details of living systems.All credit and glory go to the lab. “分子生物学家都坚信，所有的答案都会随着对特殊分子的了解而获得。”考夫曼说。 “大家都极不愿意去研究生物系统是如何运作的。”比如，基因网络研究中的吸引子的概念对他们来说是很浮夸的。 在神经科学和进化生物学领域，排斥理论的气氛要略微淡一些。但即使在这些领域，考夫曼的网络概念也被认为有点儿怪诞。他谈论庞大网络中的秩序和统计行为，却无法用这个分子或那个分子来举例说明。许多研究人员很难理解他在说些什么。“当初有人对我的基因网络的研究工作有所反应，沃丁顿就赞赏我的想法，还有其他许多人也都赞赏我的想法。这就是我得到了我的第一份工作的原因。我为此感到非常高兴，非常骄傲。但在此之后就沉寂了下来，从七十年代初期开始走下坡路。人们不再特别关心这件事了。” 考夫曼把大量时间投入到学习如何做生物实验之上了。他感到一种与他当初从哲学转向医学院时同样的冲动：他不信任自己的辩才和理论倾向。“部分原因在于我觉得自己需要做脚踏实地的工作。但另一部分的原因是我真的想知道世界究竟是如何运作的。” 考夫曼特意把研究重点放在小小的果蝇身上。基因学家在二十世纪初已对果蝇做了大量的研究。果蝇现在仍然是生物学家从事发展进程方面研究时最喜欢的实验对象。果蝇有许多有利于做实验的特点，它会出现古怪的变化，新孵出的果蝇的腿会长在触须应该长的地方，或它的生殖器会长在应该是它的头部所在的位置，等等。果蝇的这些变化给了考夫曼研究遗传形式的充分余地，他可以由此思考发育中的胚胎是如何进行自组织的。 1973 年，他对果蝇的研究使他来到了华盛顿郊外的美国健康研究所。那个研究所给他的两年任期使他得以逃脱去越南战场服兵役。 「他在华盛顿的时候已经设法推延了四年服兵役。根据众所周知的“贝雷计划”（Berry Plan），物理学家在做医学实验期间可以推延服现役。］1975年，他对果蝇的研究使他获得了宾夕法尼亚大学终身教授的职位。他开玩笑说：“我选择宾州大学，是因为那附近有非常好的印度餐厅。 ”但认真地说，他选择宾州，是因为他觉得无法再回到芝加哥大学地区的海德公园街区，把家安在那里，尽管当时加州大学也允诺了他同样的职位。他说：“那儿的犯罪率太高，种族关系太紧张了。你会感到你无力对此做出任何改善。 " 考夫曼当然不会后悔他耗费在果蝇上的时间。他发表的关于果蝇发育的论述就像他对网络形式的论述那样充满激情。但他同时也记得1982年那生动的一幕。“我在塞拉利昂山上，忽然意识到，我已经有好几年对果蝇提不出什么新的见解了。我忙忙碌碌地做着核移植实验、无性系实验和其它实验，但我其实并没有产生任何新的想法。我感到一种全面的困顿。” 不知为什么，他当时立刻就明白，是回到他起初关于基因网络和自动催化研究去的时候了。见鬼，如果没有其它内容的话，他觉得他已经善尽其职了。“我已经有权去思考我想去思考的问题了。在读完医学院、做过医生、接生过六十个婴儿、为新生儿吸抽过骨髓、照料过贲门抑制等一个年轻的医生应该做的一切事情以后，在主持过实验室，学会了如何使用闪烁计数器、如何拿果蝇做遗传实验等其它一切事情以后，尽管生物学界依然蔑视理论，但我已经有权做任何我想做的事情了。我已经满足我在牛津读书时的渴望，已经不怕自己会才思缥缈了。我现在深信自己是个优秀的理论家。这不一定是说，我总是对的。但我信任自己。” 特别是他认识到，现在是回到自动催化组的研究上去的时候了，而且这次要把它做对。他说，在1971年，他真正拥有的只不过是非常简单的计算机模拟。“我非常清楚地知道，随着方案中蛋白质数额的增加，它们之间相互反应的次数增加得会更快。所以当这个系统变得足够复杂时，就会产生自动催化现象。但在分析工作中我并没有得到多少结果。” 所以他重又开始进行计算，就像以往一样，一路研究下来，总是以发明数学公式而告终，“1983年，我耗费了整个秋季，从十月份一直到圣诞节，一直在证明各种数学定律。”他说。聚合物的数量、相互作用的次数、聚合物的接触催化反应的次数、这个巨大的反应图示中的相变次数，从中探测究竟在什么样的条件下自动催化才会发生。他怎么能证明会发生自动催化现象呢？他记得整个结果匆匆忙忙地形成于11月份，在他从印度开会回来的二十四小时的飞机航班上，“我返回到费拉德尔菲亚时简直累坍了。”他说。圣诞节那一天他匆匆忙忙地草涂下这个定律，到了1984年元旦前，他获得了结论：他在1971年只能推测，不能证明的滑稽的相变得到了确凿的证明，这个定律表明，如果化学反应过于简单，相互作用的复杂程度过低的话，相变的现象就不会发生，这个系统就会是一个低于临界点的系统。但如果相互作用的复杂性达到了一定的程度，考夫曼的数学定律就可以精确地界定这意味着什么——这个系统就会是超越临界点的系统，自动催化现象就会变得不可避免，秩序就会自由存在于其中。 真是太妙了。很显然，他接下来要做的事，就是要用更加先进的计算机模拟技术来证明这些理论设想。他说：“我已经有了关于超越临界点和在临界点之下的两种系统的设想，我急切地希望看到计算机是否能模拟这两种系统的表现。”但同样重要的是要将象征真正化学和热动力学的某种情形也编进模拟模式里去。一个更现实的模式至少可以在如何在实验室创造一个自动催化组这一方面给实验者提供指导。 考夫曼知道有两个人可以帮助他，其中的一个他已经见到过了。在巴伐利亚开会期间，他结识了罗沙拉莫斯物理学家多伊恩·法默（Doyne Farmer）。法默的想象力就和考夫曼一样丰富，一样充满活力，而且也像考夫曼一样着迷于自组织的概念。他们俩非常愉快地在阿尔卑斯山滑了一整天的雪，一直在讨论网络和自组织。他们相处得好极了，法默甚至安排考夫曼作为顾问和讲师来罗沙拉莫斯做阶段性学术访问。不久，法默又介绍考夫曼认识了伊利诺斯大学年轻的计算机专家诺曼·派卡德。 法默和派卡德自七十年代末在桑塔·克鲁兹读加州大学物理系研究生成为同学开始就一直合作默契。在加州大学读书时他们俩就都是自喻的“动力系统团体”的成员。这个团体的成员是一小群致力于那时的前沿领域——非线性动力学和混沌理论研究的研究生。这个团体的成员对非线性动力学和混沌理论作出了富有创意的贡献。这使他们的动力系统团体在詹姆士·格莱克（James Gleick）的著作《混沌》（Chaos）中占有一个篇章。《混沌》一书发表于1987年秋，差不多就在阿瑟、考夫曼和其他人为参加经济学会议而聚集桑塔费的那段时间。 当考夫曼八十年代初第一次见到法默和派卡德时，他们俩已经开始厌倦混沌理论了。 就像法默所说：“又怎么样呢？混沌的基本理论已经血肉丰满了。”他向往走在前沿的激情。在科学的前沿，事情还没有能够被完全理解。对派卡德而言，他希望自己搅到真正的复杂之中去。混沌动力学是复杂现象，当然，想想一片树叶在徐徐微风中随意摇曳吧。但这种复杂太简单化了。在树叶摇曳的情形中，只存在来自于风的一组动力。这组动力可以被一组数学等式描述出来。而这个系统只是盲目地、永远地遵循这些方程式运作。没有任何变化，也没有任何改进。“我希望超越这个，深入到生物学和心智这类更复杂、更丰富的形式之中。”派卡德说。他和法默一直在寻找切入要害的研究课题。所以当考夫曼建议，他们可以在计算机模拟自动催化系统方面相互合作时，他们便一拍即合，立即决定做这个尝试。 1985 年，当考夫曼从巴黎和耶路撒冷体完年假回来后，他们就着手这项研究。“我们之间开始了密切的合作。”考夫曼说。对化学反应的随机网络的讨论是一回事，这样的网络可以用纯数学语言来描述。但用相对真实的化学来模拟这些反应又是另一回事了，这时事情很快就变得复杂化了。 考夫曼、法默和派卡德最后得出的是一个聚合物化学的简化模型。基本的化学建设砖块，也就是根据米勒－尤瑞过程初理在初始原汤中可能形成的氨基酸和其他这类简单的混合物，在计算机模拟中用a、b、c这样的象征性符号来表达。这些建设砖块能够相互连接成链，形成更大的分子，比如像accddbacd。这些更大的分子反过来又会发生两种化学反应。它们可能分裂开来： accddbacd→accd＋dbacd或者它们也可以反过来，最终合为一体：bbacd＋cccba→bbcadcccba 每一组反应都有一个相关的数——化学家将其称为率常数——这个数能够决定在没有接触剂的情况下发生化学反应的频率。 当然，这个实验的全部意义在于观察在有接触剂的情况下会发生什么情况。所以考夫曼、法默和派卡德必须找到能够分辨哪一个分子触发哪一种化学反应的方法。他们尝试了好几种方法，其中考夫曼提议采取的一种与其他方法的效果类似的方法，即只是选取一系列的分子，比如像abccd，然后任意指定每一个分子的化学反应，比如baba＋ccda→babaccda。 在进行模拟时，一旦所有的反应速率和催化强度被界定清楚后，考夫曼、法默和派卡德就让计算机开始丰富他们模拟的原始池塘，源源不断地提供像a、b、aa这一类的分子“食物”。然后他们就坐下来，看看他们的模拟会产生什么样的结果。 在很长时间内，他们的模拟产生不出多少结果。这很令他们沮丧，但却并不令人吃惊。反应速率、催化强度和食物供给率，所有这些参数都可能有误。要做的是改变这些参数，然后再看看什么参数会起作用，什么参数不会起作用。他们在这样做时偶然发现，当他们把参数修改到有利的范围之内时，模拟的自动催化组就产生了。更进一步的是，自动催化组形成的条件，似乎正像考夫曼在他的关于抽象的网络定理中所预测的那样。 1986年，考夫曼和他的合作者发表了他们的研究结果。虽然这时法默吸收了一名研究生，里查德·巴格雷（Richard Bayley），来拓展和大幅度加快这个模拟实验，但法默和派卡德这时早已兴趣别移了。考夫曼自己也开始进一步思考进化中自组织情形发生的其他方式。但在这次计算机模拟实验之后，他比以往更深刻地感到，他已经真正面对造物主的奥秘了。 他记得有一次独自登上泰后湖畔的塞拉斯山，到他最爱去的豪塞泰尔瀑布。他回忆说，那是一个怡人的夏日。他坐在瀑布旁的一块石头上，思考着自动催化的问题及其意义。“我突然明白了，上帝已经部分地向我展示了宇宙运作的奥秘。”当然，他指的不是通常人们认为的那个上帝，考夫曼从来没有相信过有上帝的存在。“但我有了一种理解宇宙的神圣感觉，一个展现在我面前的宇宙，一个我身为其中一部分的宇宙。事实上，这是与虚荣自负正好相反的一种感觉。我感觉到上帝正在向愿意倾听的人展示世界运行的奥秘。” 他说：“这是一个美好的时刻，一个我离宗教体验最接近的时刻。”