Mediamatic Magazine Vol. 9#1 Hein Masseling 1 Jan 1998

Kauffman

Origins of Order

If you ask questions on the origin of species, ecosystems or other complex biological structures, the answers will undoubtedly include the term 'natural selection'. Natural selection, as formulated by Darwin, still holds a central place in the modern theory of evolution. But although the importance of natural selection is virtually undisputed within the field of biology, at the same time it is generally assumed that other mechanisms could also play an important role in the process of evolution. One such mechanism, which recently has been attracting a great deal of attention, is self-organization.

Stuart Kauffman's belief that self-organization plays an important role in biological processes dates back to his early student days. Kauffman, whose current activities include an appointment at the University of Santa Fé - which is pivotal in the study of Artificial Life - has given shape to this conviction over the last few decades. Although, at the start of his studies he was virtually alone in his opinion, in recent years his ideas have enjoyed more and more widespread attention and recognition. In two recently published books, Kauffman reports on his research, which is mainly based on computer simulation. Origins of Order is a seven-hundred page book, which, because of its massive form, appears to be written mainly for the profession. At Home in the Universe, however, embraces a wider public. Kauffman tries to show how self-organization can play a role in evolution and its interaction with natural selection.

At first sight, it seems strange to make a distinction between the concepts of 'natural selection' and 'self-organization', because natural selection is, in fact, an excellent example of self-organization. It is this mechanism which makes the guiding hand of God, or an underlying rational plan, superfluous in the explanation of the perceived order. Kauffman, however, gives a restricted meaning to self-organisation. As he formulates it, natural selection is a historical process. A selection is made from the variation which is created 'by accident' (mutation, recombination, floods, meteorites, etc.). Natural selection is an opportunistic process. If evolution were to repeat itself, it is more than likely that the outcome would not be the same as the present result. For example, the chance that certain mutations would occur twice in the same way is inexpressibly small. Natural selection is chance caught on the wing, as Jacques Monod, the French geneticist, phrased it. We, human beings, can count ourselves lucky we exist at all. According to Kauffman, the emphasis which, since Darwin, has been placed on natural selection as a mechanism of evolution, has pushed evolutionist biology too far into the field of historical science. In his opinion, what we have lost sight of is that certain biological processes take place inevitably, just as physical and chemical processes inevitably take place. Biological processes can also be a-historical. Kauffman reserves the term 'self-organization' for such a-historical biological processes. He is convinced of the importance of such processes in evolution. Perhaps we humans need not be so surprised at having come into existence during the course of evolution. We the expected, we at home in the universe, as Kauffman formulates it.

According to Kauffman, one of the processes in which self-organization has probably played a role is the emergence of life. In the pre-biotic environment on Earth, there was probably an ever growing diversity of organic molecules, which, moreover, also became more and more complex. At a given moment, according to Kauffman, there must have been a critical diversity of molecules and interactions between them, which enabled a new phenomenon to emerge: a self-preserving network of organic molecules and their interactions, with reactions being catalyzed by reaction products of the network itself. In short, a certain minimum complexity must have generated life. In the same way that a substance will crystallize in certain conditions, so life will crystallize given sufficient complexity. Literally. It is not surprising, therefore, in Kauffman's opinion, that the most elementary organisms, the bacteria, are still relatively complex. According to him, the emergence of life does not depend on a specific situation of origin. With a different composition of the pre-biotic environment, life would probably have taken on a different shape, but there is no doubt in Kauffman's mind that it would have emerged. Nor that, in the near future, humanity will be able to generate life once again...

Another process in which, according to Kauffman, self-organization plays an important role, is cell differentiation. The genes of an organism constitute in fact a network of genes which can each be in one of two states; active (the gene is expressive) or non-active (the gene is not expressive). It is generally true of all cells of an organism that they contain the same genes, and that the pattern of active genes determines into which type of cell (neuron, blood corpuscule, muscle cell, etc.) a certain cell will develop. However, the number of cell types making up an organism is much smaller than the number of possible patterns of active and non-active genes. Human beings have some 100,000 genes, and the number of possible patterns is therefore unimaginably great: 2 to the power of 100,000. Such a number of possibilities is of no use from a biological point of view. However, Kauffman's computer model (broadly speaking: a network consisting of mutually connected elements which can be in either of two states) shows that the number of possible patterns is largely determined by the parameter expressing the number of interconnections between elements. Genes can, for example, influence each other by means of repressor molecules. If, in the computer model, each gene is directly influenced by two other genes, which is a biologically sound assumption, the number of possible conditions of the genome turns out to be reduced to approximately the root of the number of elements in the network. For human beings, this would mean that the genetic network allows for some 317 different conditions. In view of the number of possible conditions and the simplifications in Kauffman's model, this corresponds reasonably well with the 256 types of cells from which man is built up. The number of types of cells predicted by computer simulation turns out to correspond equally well with the reality of a great diversity of other organisms. Order, biologically plausible order, says Kauffman, emerges spontaneously, it crystallizes at a certain parameter value. He calls this Order for free.

To Kauffman, the main question is that of the relationship between self-organization and natural selection. Self-organization can, by its spontaneous, crystallizing nature, as it were consolidate a situation which has been brought about by natural selection. But Kauffman suspects that self-organization could play an even more vital role. He shows that natural selection is not always able to resist strong, species-unfriendly, mutative pressure, and could, in its opportunism, encumber an organism with a no more than mediocre kind of fitness. In such cases, the possibilities for the organism, and therefore the species, to adjust after all, become rather limited. Kauffman expects (hopes?) that self-organization, through its a-historical, spontaneous, nature, can guarantee sufficient stability for the structures (cells, genomes) in whose emergence it plays a role, to become less vulnerable to the characteristic opportunism of natural selection. He proposes that self-organization could possibly generate the structures which could serve as the basis for further development by means of natural selection. Slightly confusingly, Kauffman at the same time assumes that the circumstances in which self-organization can occur are brought about by natural selection. Natural selection, in a manner of speaking, fine-tunes the parameters of the system and, at a given value of these parameters, the system organizes itself, as with the genome, into a biologically effective system. Of course this, in turn, raises the question of whether self-organization, due to its dependence on natural selection, is not largely historically determined after all. We the expected, we at home in the universe would seem to be a somewhat over-enthusiastic catch phrase when specifically referring to the human species.

Kauffman himself admits that his scenario is rather speculative. But it is a challenging scenario, and it has given the theory of evolution an important new impulse. Origins of Order is a book which is often referred to. However, what I found most remarkable about Kauffman's books was not the idea of self-organization itself. Indeed, it is a rather self-evident idea that even biological processes can have the inevitable character of physical processes, even though, in evolutionist biology, this idea has, until recently, remained hidden behind the concept of natural selection as the explanatory mechanism. It is another phenomenon, which in fact unexpectedly presented itself in his computer simulations, that impressed me the most. Kauffman calls this the edge of chaos, the area between stability and chaos. Stable systems are not very sensitive to disturbing influences. Chaotic systems however, are extremely sensitive to differences in the situation of origin. For example, in the computer models which Kauffman used to simulate the interaction between genes of the genome, or between the various species within an eco-system, it turned out that the system could either be stable or chaotic, depending on the value of certain parameters (such as the number of interrelations between genes or species). However, at certain parameter values, the system hovered between both extremes:// the edge of chaos. This is the area where the ability to evolve, the evolvability, turns out to be the greatest. This is easy to imagine: an evolving system can benefit from stability (it needs to be strong) as well as from flexibility (it must be able to adjust). The border between stability and chaos provides precisely this compromise. Miraculously, the evolving systems of Kauffman's simulations turned out to assume precisely the parameter values which brought the system to this border between stability and chaos. The blind process of evolution appeared capable of finding the spot where the possibilities for further evolution are greatest! An amazing outcome indeed. According to Kauffman, once again, it is natural selection which fine-tunes the parameters of an evolving system in such a way that this system ends up on the border between order and chaos. Of course, it is a question of whether natural selection is indeed an opportunistic process, as Kauffman assumes it is. If this is the case then the process directs an evolving system towards the spot which offers this evolving system the best //perspectives.

Kauffman is convinced of the 'big picture': he believes that important and non-coincidental parallels can be drawn between self-organization and the themes around the 'edge of chaos' in biological processes, and all manner of cultural and technological developments. Not only do these processes resemble each other, they are actually the same. He refers, for example, to the similarities between the emergence of new technological developments and the emergence of life. According to Kauffman, both depend on minimal complexity. And he compares democracy with a system existing on the border of order and chaos. In the same way that the area on the border of order and chaos provides the best possibilities for further evolution, so does democracy provide the best perspective on a solution which is acceptable to as many people as possible, or is at least bearable. Here, however, Kauffman\\\\\\\\\\\\\\\\\\\\\\\\\\\\'s discourse gives us little to go on, simply because there has been little to no specific research in this area. He readily admits to this, but proposes that this is all the more reason to start such research soon.

'Self-organization' and 'the edge of chaos': new terms which could possibly redefine the theory of evolution, biology in general, and, who knows, sociological and economic models as well. In Kauffman, they have found an enthusiastic advocate. And his enthusiasm is sincere: you can picture him at his computer, grinning from ear to ear at the screen where everything is happening just as he always hoped it would.

^^

At Home in the Universe - The Search for the Laws of Self-Organization and Complexity

Oxford 1995, isbn 0 19 509599 5, English text, 329 pp, $ 16.95^^

translation OLIVIER / WYLIE

If you ask questions on the origin of species, ecosystems or other complex biological structures, the answers will undoubtedly include the term 'natural selection'. Natural selection, as formulated by Darwin, still holds a central place in the modern theory of evolution. But although the importance of natural selection is virtually undisputed within the field of biology, at the same time it is generally assumed that other mechanisms could also play an important role in the process of evolution. One such mechanism, which recently has been attracting a great deal of attention, is self-organization.

Stuart Kauffman's belief that self-organization plays an important role in biological processes dates back to his early student days. Kauffman, whose current activities include an appointment at the University of Santa Fé - which is pivotal in the study of Artificial Life - has given shape to this conviction over the last few decades. Although, at the start of his studies he was virtually alone in his opinion, in recent years his ideas have enjoyed more and more widespread attention and recognition. In two recently published books, Kauffman reports on his research, which is mainly based on computer simulation. Origins of Order is a seven-hundred page book, which, because of its massive form, appears to be written mainly for the profession. At Home in the Universe, however, embraces a wider public. Kauffman tries to show how self-organization can play a role in evolution and its interaction with natural selection.

At first sight, it seems strange to make a distinction between the concepts of 'natural selection' and 'self-organization', because natural selection is, in fact, an excellent example of self-organization. It is this mechanism which makes the guiding hand of God, or an underlying rational plan, superfluous in the explanation of the perceived order. Kauffman, however, gives a restricted meaning to self-organisation. As he formulates it, natural selection is a historical process. A selection is made from the variation which is created 'by accident' (mutation, recombination, floods, meteorites, etc.). Natural selection is an opportunistic process. If evolution were to repeat itself, it is more than likely that the outcome would not be the same as the present result. For example, the chance that certain mutations would occur twice in the same way is inexpressibly small. Natural selection is chance caught on the wing, as Jacques Monod, the French geneticist, phrased it. We, human beings, can count ourselves lucky we exist at all. According to Kauffman, the emphasis which, since Darwin, has been placed on natural selection as a mechanism of evolution, has pushed evolutionist biology too far into the field of historical science. In his opinion, what we have lost sight of is that certain biological processes take place inevitably, just as physical and chemical processes inevitably take place. Biological processes can also be a-historical. Kauffman reserves the term 'self-organization' for such a-historical biological processes. He is convinced of the importance of such processes in evolution. Perhaps we humans need not be so surprised at having come into existence during the course of evolution. We the expected, we at home in the universe, as Kauffman formulates it.

According to Kauffman, one of the processes in which self-organization has probably played a role is the emergence of life. In the pre-biotic environment on Earth, there was probably an ever growing diversity of organic molecules, which, moreover, also became more and more complex. At a given moment, according to Kauffman, there must have been a critical diversity of molecules and interactions between them, which enabled a new phenomenon to emerge: a self-preserving network of organic molecules and their interactions, with reactions being catalyzed by reaction products of the network itself. In short, a certain minimum complexity must have generated life. In the same way that a substance will crystallize in certain conditions, so life will crystallize given sufficient complexity. Literally. It is not surprising, therefore, in Kauffman's opinion, that the most elementary organisms, the bacteria, are still relatively complex. According to him, the emergence of life does not depend on a specific situation of origin. With a different composition of the pre-biotic environment, life would probably have taken on a different shape, but there is no doubt in Kauffman's mind that it would have emerged. Nor that, in the near future, humanity will be able to generate life once again...

Another process in which, according to Kauffman, self-organization plays an important role, is cell differentiation. The genes of an organism constitute in fact a network of genes which can each be in one of two states; active (the gene is expressive) or non-active (the gene is not expressive). It is generally true of all cells of an organism that they contain the same genes, and that the pattern of active genes determines into which type of cell (neuron, blood corpuscule, muscle cell, etc.) a certain cell will develop. However, the number of cell types making up an organism is much smaller than the number of possible patterns of active and non-active genes. Human beings have some 100,000 genes, and the number of possible patterns is therefore unimaginably great: 2 to the power of 100,000. Such a number of possibilities is of no use from a biological point of view. However, Kauffman's computer model (broadly speaking: a network consisting of mutually connected elements which can be in either of two states) shows that the number of possible patterns is largely determined by the parameter expressing the number of interconnections between elements. Genes can, for example, influence each other by means of repressor molecules. If, in the computer model, each gene is directly influenced by two other genes, which is a biologically sound assumption, the number of possible conditions of the genome turns out to be reduced to approximately the root of the number of elements in the network. For human beings, this would mean that the genetic network allows for some 317 different conditions. In view of the number of possible conditions and the simplifications in Kauffman's model, this corresponds reasonably well with the 256 types of cells from which man is built up. The number of types of cells predicted by computer simulation turns out to correspond equally well with the reality of a great diversity of other organisms. Order, biologically plausible order, says Kauffman, emerges spontaneously, it crystallizes at a certain parameter value. He calls this Order for free.

To Kauffman, the main question is that of the relationship between self-organization and natural selection. Self-organization can, by its spontaneous, crystallizing nature, as it were consolidate a situation which has been brought about by natural selection. But Kauffman suspects that self-organization could play an even more vital role. He shows that natural selection is not always able to resist strong, species-unfriendly, mutative pressure, and could, in its opportunism, encumber an organism with a no more than mediocre kind of fitness. In such cases, the possibilities for the organism, and therefore the species, to adjust after all, become rather limited. Kauffman expects (hopes?) that self-organization, through its a-historical, spontaneous, nature, can guarantee sufficient stability for the structures (cells, genomes) in whose emergence it plays a role, to become less vulnerable to the characteristic opportunism of natural selection. He proposes that self-organization could possibly generate the structures which could serve as the basis for further development by means of natural selection. Slightly confusingly, Kauffman at the same time assumes that the circumstances in which self-organization can occur are brought about by natural selection. Natural selection, in a manner of speaking, fine-tunes the parameters of the system and, at a given value of these parameters, the system organizes itself, as with the genome, into a biologically effective system. Of course this, in turn, raises the question of whether self-organization, due to its dependence on natural selection, is not largely historically determined after all. We the expected, we at home in the universe would seem to be a somewhat over-enthusiastic catch phrase when specifically referring to the human species.

Kauffman himself admits that his scenario is rather speculative. But it is a challenging scenario, and it has given the theory of evolution an important new impulse. Origins of Order is a book which is often referred to. However, what I found most remarkable about Kauffman's books was not the idea of self-organization itself. Indeed, it is a rather self-evident idea that even biological processes can have the inevitable character of physical processes, even though, in evolutionist biology, this idea has, until recently, remained hidden behind the concept of natural selection as the explanatory mechanism. It is another phenomenon, which in fact unexpectedly presented itself in his computer simulations, that impressed me the most. Kauffman calls this the edge of chaos, the area between stability and chaos. Stable systems are not very sensitive to disturbing influences. Chaotic systems however, are extremely sensitive to differences in the situation of origin. For example, in the computer models which Kauffman used to simulate the interaction between genes of the genome, or between the various species within an eco-system, it turned out that the system could either be stable or chaotic, depending on the value of certain parameters (such as the number of interrelations between genes or species). However, at certain parameter values, the system hovered between both extremes:// the edge of chaos. This is the area where the ability to evolve, the evolvability, turns out to be the greatest. This is easy to imagine: an evolving system can benefit from stability (it needs to be strong) as well as from flexibility (it must be able to adjust). The border between stability and chaos provides precisely this compromise. Miraculously, the evolving systems of Kauffman's simulations turned out to assume precisely the parameter values which brought the system to this border between stability and chaos. The blind process of evolution appeared capable of finding the spot where the possibilities for further evolution are greatest! An amazing outcome indeed. According to Kauffman, once again, it is natural selection which fine-tunes the parameters of an evolving system in such a way that this system ends up on the border between order and chaos. Of course, it is a question of whether natural selection is indeed an opportunistic process, as Kauffman assumes it is. If this is the case then the process directs an evolving system towards the spot which offers this evolving system the best //perspectives.

Kauffman is convinced of the 'big picture': he believes that important and non-coincidental parallels can be drawn between self-organization and the themes around the 'edge of chaos' in biological processes, and all manner of cultural and technological developments. Not only do these processes resemble each other, they are actually the same. He refers, for example, to the similarities between the emergence of new technological developments and the emergence of life. According to Kauffman, both depend on minimal complexity. And he compares democracy with a system existing on the border of order and chaos. In the same way that the area on the border of order and chaos provides the best possibilities for further evolution, so does democracy provide the best perspective on a solution which is acceptable to as many people as possible, or is at least bearable. Here, however, Kauffman's discourse gives us little to go on, simply because there has been little to no specific research in this area. He readily admits to this, but proposes that this is all the more reason to start such research soon.

'Self-organization' and 'the edge of chaos': new terms which could possibly redefine the theory of evolution, biology in general, and, who knows, sociological and economic models as well. In Kauffman, they have found an enthusiastic advocate. And his enthusiasm is sincere: you can picture him at his computer, grinning from ear to ear at the screen where everything is happening just as he always hoped it would.

^^

At Home in the Universe - The Search for the Laws of Self-Organization and Complexity

Oxford 1995, isbn 0 19 509599 5, English text, 329 pp, $ 16.95^^

translation OLIVIER / WYLIE