Home
Was
Darwin
Wrong?		Home | Intro | About | Feedback | Prev | Next | Search
Kauffman at home in the Universe.

The secret of life is auto-catalysis
review by Gert Korthof
20 Oct 1998 (updated 5 Mar 2009)

  Kauffman's autocatalytic set theory is a revolutionary theory about the origin of life. It embodies a revolutionary new way of thinking about life. We could compare what Kauffman did for the theory of evolution, with what Mendel did for genetics. Mendel introduced mathematical methods and postulated abstract 'factors' at a time that genes, DNA and chromosomes were unknown. Nothing was known about the material basis of heredity. Naturally scientists at the time didn't like those abstract hereditary 'factors' and didn't know what to think about 'factors' that were transmitted according to mathematical rules such as 1 : 3 ratio or even ratio's like 9:3:3:1. If Mendel had those 'factors' in a test-tube, he would be believed immediately. His work needed to be rediscovered and eventually the science of genetics grew out of it. Kauffman used mathematics to come to grips with the problem of the origins of order and developed abstract autocatalytic sets, genetic networks and fitness landscapes. Again people hesitate or reject it.
    In this review I start with (§1) a very short description of autocatalytic sets, go on to describe shortly five of Kauffman's criticisms of Darwinism:
§2: natural selection cannot be the sole source of order,
§3: the accidental and historical nature of Darwinism,
§4: the reductionistic nature of Darwinism,
§5: the gene and DNA centred view of Neo-Darwinism,
§6: can gradual accumulation of minor improvements work ?
Then I will give (§7) evidence for the theory (§8) and some weak points of Kauffman's theory as I see them and finally (§9) a few words about what creationists said about Kauffman. A lot will not be discussed. Yet, in the end I hope it will be clear why I placed Kauffman's At Home in the Universe in the category "non-orthodox-Darwinism". Although the book was written for a wider public, it is not an easy read, but absolutely worth the effort.
book "At Home in the Universe.
The Search for Laws of Self-Organization and Complexity."
by Stuart Kauffman.
Viking
1995
321 pages

Contents:
1. At Home in the Universe  3
2. The Origins of Life 31
3. We the Expected 47
4. Order for Free 71
5. The Mystery of Ontogeny 93
6. Noah's Vessel113
7. The Promised Land131
8. High-Country Adventures149
9. Organisms and Artifacts191
10. An Hour upon the Stage207
11. In Search of Excellence245
12. An Emerging Global Civilization273
Bibliography305
Index307-321

1
Autocatalytic sets 		'Autocatalytic set theory' is a theory about the origin of life (19). Autocatalytic set theory is an absolute simple model with absolutely non-trivial properties. It is a very useful model because it is implemented in a computer program (the model is executable). The idea behind the model is that life is a collection of molecules catalysing each others formation. (In §4 more about reductionism!). And further that every molecule (peptide, protein) has the capacity to catalyse some reaction. The model enables us to explore what happens when molecules in a prebiotic chemical mixture are catalysing each others formation. This is impossible without the help of a computer program. The model is without knowledge of which molecules need to be present in the mixture, and without knowledge which specific molecule is catalysing which reaction. An important assumption is that, if enough different molecules are present, molecules will catalyse the formation of other molecules by chance. Thereby becoming members of the set. In a collectively auto-catalytic set the molecules speed up the very reactions by which they themselves are formed. Now, if there are enough different molecules and assuming a fixed probability that any reaction in the mixture is catalysed by at least one molecule in the mixture, the model demonstrates that the mixture will suddenly transform into a collectively autocatalytic whole. In other words: there is a threshold. The formation of every member is catalysed by at least one other member. And so the set as a whole is stable. Translated into 'real-life': assuming enough food and energy, high enough concentrations and a fixed probability for chance catalysis, the mixture of molecules can transform into a stable and self-sustaining set. In other words: will be alive! So, what we have here is nothing less than a new and revolutionary theory of the origin of life from a random mixture of chemicals without the need for DNA and natural selection! And at the same time it introduces a completely new source of order in biology. A deeper source of order than the specifics of current metabolism. I'll return to this later.  
2
Neo-Darwinism is incomplete		 Neo-Darwinism is not enough to explain order of the living world: cells, organisms, ecosystems. Much of the order in organisms may not be the result of selection at all, but of the spontaneous order of self-organized systems (autocatalytic sets). So the combination of Natural Selection and Mutation cannot be the sole source of order. The role of Natural Selection is that it acts upon natural order. Kauffman is not anti-Darwinist in the sense that he rejects the existence of natural selection. It does play an important role in evolution, but he maintains that it cannot be the sole source of order. "We stand in the need of a new conceptual framework that allow us to understand an evolutionary process in which self-organization, selection and historical accident find their natural places with one another." (p150). So it's more appropriate to describe Kauffman as opposed to ultra-Darwinism [1]. Kauffman's criticism is motivated by his knowledge of other sources of order, Darwinists and biologists didn't know about.  
3
Biology has become a deeply historical science		 Since Darwin, characteristics of organisms are explained by common descent, and common descent implies almost always historical accidents. Because in neo-Darwinism the existence and the evolution of life depends so much on random events, life itself is seen as an accident. An improbable accident. "We are not supposed to be here !" (p43). "Biology has come to seem a science of the accidental" (p7). "Many features of organisms are not merely historical accidents." (p25). Biology need not be a historical science, according to Kauffman. He tries to find natural laws. Laws of complexity. Kauffman: "We are no accidents, but a natural law created us". "If life were bound to arise, not as an incalculably improbable accident, but as an expected fulfilment of he natural order, then we truly are at home in the universe." (p20). This explains the title of his book. It is clear that if the origin of life is an improbable event, we are not going to have a convincing scientific explanation. However, if we can find laws, we could explain the origin of life.
   Please note that Kauffman did not eliminate randomness. Kauffman's theory of autocatalytic sets is based on the probability that a randomly chosen protein catalyses a randomly chosen reaction (p147). And further: the fortunate set of proteins becoming an autocatalytic set, is a random subset of all possible autocatalytic sets [2]. And in so far as I understand it, the particular metabolism of the first living cells, must have been a frozen accident and there are millions of other possible metabolisms life could use. However, the theory of autocatalytic sets is a law, albeit a statistical law. 		 
4
Reductionistic nature of neo-Darwinism 		Kauffman's answer to reductionism is: holism. Kauffman's holism is directed against the gradualism and gene and DNA centered view neo-Darwinism. Holism, because the whole has properties that parts do not have: the emergent properties. This is a non-mystical holism. Holism means that if complexity increases beyond a threshold, life emerges suddenly as a whole, not simple but complex from the start and not gradual, but suddenly (p24,47,69).
If all properties of living systems depend on every detail of their structure, then there is no hope of understanding living systems. Life is not to be located in its parts, but in the collective emergent properties of the whole.
 		 
5
Gene-centered nature of neo-Darwinism 		"Life does not depend on the magic of Watson-Crick base pairing or any other specific template-replicating machinery. Life lies ... in the property of catalytic closure among a collection of molecular species". (p50) The origin of life cannot be explained if DNA would be essential (p72) and autocatalytic sets can evolve without a genome. Life started without DNA ! Life started as a collectively autocatalytic set of molecules. If one could create autocatalytic sets in the laboratory, one could create life! (p147). Kauffman's anti-DNA view holds for the origin of life, but in the ontogeny of multicellular organisms networks of genes are needed to enable cell differentiation. Kauffman's view of life is opposed to Hubert Yockey's view of life. Yockey [3] considers information as the secret of life. Information coded in DNA. Two opposing views of life.  
6
Can gradualism work? 		"One of the most important presuppositions of Darwin's entire thesis is gradualism, the idea that mutations to the genome can cause minor variations in the organism's properties, which can be accumulated piecemeal, bit by bit, over the eons to create the complex order found in the organisms we observe." (p151).
This criticism sounds familiar to those who read Denton, Behe, or virtually any creationist. Kauffman has his own reasons to doubt gradualism: the complexity paradigm. Can all organisms be "improved"; and ultimately assembled by accumulating a succession of minor modifications ? Can we extrapolate from modifying certain characteristics by artificial selection to all characteristics of all organisms in nature ? According to Kauffman, Darwin almost certainly was wrong, because in some [4] complex systems any minor mutation causes catastrophic changes in the system, in stead off improvements. Kauffman illustrates this with mathematical examples and tries to characterize what kinds of complex systems can be build by an evolutionary random search in a reasonable time.
Gradualism is closely connected with reductionism. If an organism is the sum of relatively independent parts, it will be easy to mutate the parts stepwise without adverse effects on the whole. However, when every component of an organism is strongly connected to all other components, a minor mutation in one component influences all other components. 		 
7
Evidence for the theory 		The revolutionary character of Kauffman's autocatalytic sets will become clear when we compare it with the failed attempts of Robert Shapiro [5] and Hoyle & Wickramasinghe. They calculated the chance of random assembly of an enzyme of 200 amino acids long and a bacterium with 2000 enzymes. On the basis of these calculations life could simply not have started. But since we are here, something must be wrong with the argument! Kauffman's solution is this:
"It is not necessary that a specific set of 2000 enzymes be assembled... Whenever a collection of chemicals contains enough different kinds of molecules, a metabolism will crystallize from the broth." (p45).
This was the birth of autocatalytic sets. The revolutionary aspect of this proposal is that nobody did make the step from the specific set of 2000 enzymes which happens to occur in the bacterium, to the set of all possible metabolic sets. I think we will never make progress in understanding the origin of life, if we don't analyse life as a subset of the set of all possible life forms. In my view it is for the first time that the right kind of probability calculus for the origin of life is introduced. As I see it, it rejects the notion that only the metabolism we find today in cells, is uniquely capable of producing and sustaining life (against Denton[6]). Compare it with throwing dice. To calculate the probability of getting 8 by throwing two dice, we need to know all possible ways to get 8. We could not hope to obtain the right probability if we limited the possibilities to 4+4 and forget about 3+5, 2+6, etc. The thing to do is: decide if the idea is basically right and then use it, and test it and elaborate it. The outcome could be that many alternative metabolisms are in fact possible, or no alternative metabolism is possible, but to find out, we have to explore it in the first place!

Is there evidence that supports the theory of autocatalytic sets? If a theory explains brute facts, we have a useful theory. Kauffman shows in a graph that the number of cell types increases as a square-root function of the number of genes of an organism (p109). According to that law, a human, with an estimated 100,000 genes, should have about 317 cell types. The real number of celltypes appears to be 256, but the number of genes is not yet known exactly [17]. As far as I know there isn't a theory that predicts a number of celltypes at all.
I asked Stuart Kauffman by email (21 Oct 1998) whether new evidence was produced after the publication of At Home in the Universe. His response:

Gunter von Kiedrowski[15], then at U. Freiburg in Germany, several years ago published work on a collectively autocatalytic set of two DNA hexamers that mutually ligated the two pairs of DNA trimers composing the two hexamers. Meanwhile, Reza Ghadiri at the Scripps Institute in La Jolla, California has made an autocatalytic peptide, Nature August 2 years ago [9,10], and nearly collectively autocatalytic sets more recently [11,12].
Unexpected evidence comes from Lou Allamandola [13]:
"The most amazing thing is that we start with something really simple. And then suddenly we're making this enormous range of complex molecules. When I see this kind of complexity forming under these exceedingly extreme conditions, I begin to really believe that life is a cosmic imperative."
and from Biliang Zhang & Tom Cech [14], who isolated RNA's that could efficiently link specific amino acids together. These pseudo-ribosomes were selected from a random pool of 1015 synthetic RNA's. So, there is enough evidence to invalidate the claim, that complexity theory is 'fact-free science'. Recently Yao et al [16] described a four-component peptide system that is capable of auto- and cross-catalysis and which supports the suggestion that self-replicating peptides may have played a role in the origin of life. In 2009 Science reported an RNA enzyme catalysing the replication of itself (20). 		 
8
Difficulties of theory 		Finally, some difficulties of the theory of autocatalytic sets, as I perceive them: If enough diversity is present, an autocatalytic set arises. Autocatalytic sets depend on reversible reactions. However, if there is enough molecular diversity to allow a randomly chosen chemical to catalyze a randomly chosen reaction, then there should be every reason to expect that also a randomly chosen chemical will inhibit a randomly chosen reaction. That is to say irreversible binding of one molecule to another. It would prevent autocatalytic closure. The system would freeze or disintegrate. The system won't catch fire; come to life. It would prevent the origin of life. I do not doubt that Kauffman's computer model of autocatalytic sets produces the emergent properties he observes. But he leaves out toxins, inhibitors, poisons, enzym-blockers which are a fact of life (toxicology and pharmacology would not exist without chemical inhibition). Kauffman mentions inhibitors briefly in The Origins of Order[2], and concludes that they do not prevent autocatalytic closure, but inhibition is not a basic component of his model. Maybe indirect autocatalysis results from an inhibition of an inhibition? The only real life limits to the theoretical model Kauffman mentions, are supply of food and energy, and concentrations, but they are not part of the computer model! The only danger autocatalytic sets have to face, an explosion of molecular diversity, is again based on catalysis. No inhibition enters to model. An eye-opener is Kauffman's question: why do our cells digest food, breaking it apart into smaller molecules, only to build them up again? Why don't our cell's fuse with food cells? Kauffman is afraid of an explosion of molecular catalysis. Again inhibition does not enter the picture.
    Kauffman is clear about the existence of a threshold of complexity in At Home in the Universe, but not about the absolute value of the threshold. Fig 3.1 He shows a simple autocatalytic set in figure 3.1 with only 4 different molecules (A,B,AB,BA). So if a autocatalytic set is alive, a system of 4 molecules is alive? If the threshold is 4, the origin of life would be easy! It seems that the observed threshold in nature is a brute fact, not predicted by the theory of autocatalytic sets. In figure 3.4 the ratio of threads/buttons is displayed, not an absolute number of buttons. Does the threshold not depend on the absolute number of different molecules ? Yes, in The Origins of Order [2] Kauffman calculates the minimum complexity (threshold) of a closed autocatalytic set as 6165 different molecules under favourable conditions and 34 million under unfavourable conditions. It's a pity that he did not mention these results in At Home in the Universe. Because the simplest free-living organism, Mycoplasma genitalium, has only 468 genes! 		 
9
Creationists about Kauffman 		Phillip Johnson compared Neo-Darwinism with a sinking battleship:
There will be heroic efforts to save the ship, of course, and some plausible rescuers will invite the officers to take refuge in electronic lifeboats equipped with high-tech gear like autocatalytic sets and computer models of self-organizing systems. [7]
Johnson wrongly interpreted Kauffman's work as saving Darwinism. Johnson simply did not grasp Kauffman's goals: to understand the origin of order in nature. As an outsider he wasn't able to understand the goal of science: to understand nature. As a creationist and Darwin critic, he failed to detect and take advantage of Kauffman's criticism of neo-Darwinism. Finally, Johnson, occupied with his war against atheists, overlooked that Kauffman's 'We­the­Expected' philosophy is something what all religious people support.
Michael Behe did quote Kauffman's criticism of Darwinism, but his opinion about complexity theory is:
No proponent of complexity theory has yet gone into a laboratory, mixed a large variety of chemicals in a test tube, and looked to see if self-sustaining metabolic pathways spontaneously organize themselves.[8]
We know now, that there is evidence (See §7). Further Behe (he is a professor of biochemistry) complains that there are no names of specific chemicals in Kauffman's theory.
John Maynard Smith, Kauffman's old mentor, has accused him of practicing "fact-free science". [8]
This did not prevent John Maynard Smith to state a few years later that
"Autocatalytic cycles are important for the origin of life" [18]

Michael Denton notes that Kauffman's "overall conclusion is consistent with my own" [6]. Indeed, the phrases "At Home in the Universe" and "We the Expected" aptly describe the main theme of Denton's book. Denton's universe is biocentric and anthropocentric because life depends on the precise values of cosmological, physical and chemical parameters. Life depends on chemicals uniquely fit for their biological roles, and are irreplaceable by other chemicals. Let's call it 'specifics-metaphysics'. Kauffman's universe is fit for life because the probability of an arbitrary protein to catalyze another chemical reaction is high enough to guarantee autocatalysis. Both arguments point to the inevitability of life. Both views reject the accidental and historical nature of neo-Darwinism (see §3). However, thinking about the assumptions of autocatalytic sets, I realised that there could exist in fact no greater discrepancy between the two approaches. Kauffman eliminated specifics: no special or unique chemicals are needed for autocatalytic sets: any mixture of sufficient complexity will produce autocatalysis. Many metabolisms would be possible. In Denton's universe life utterly depends on 'the specific' and 'the unique', on a long chain of coincidences. There could be only one metabolism: the metabolism we have. Denton didn't notice this fundamental difference. He didn't notice either that his 'brute facts' could be a serious challenge to autocatalytic sets. Are autocatalytic sets capable of incorporating those specifics?

  Notes

  1. Niles Eldredge(1995) Reinventing Darwin. The Great Evolutionary debate. (on this site)
  2. Stuart Kauffman(1993) The Origins of Order, (p311,312). Review.
  3. Hubert Yockey(1992) Information theory and molecular biology (review on this site)
  4. that would mean that the majority is capable of stepwise improvement.
  5. Robert Shapiro(1986) Origins: A Skeptic's Guide to the Creation of Life on Earth.
  6. Michael Denton(1998) Nature's Destiny. (review on this site)
  7. Phillip Johnson(1993) Darwin on Trial, p170. (review on this site)
  8. Michael Behe(1996) Darwin's Black Box. The biochemical challenge to evolution, p190 and 156. (review on this site)
  9. David H. Lee, Juan R. Granja, Jose A. Martinez, Kay Severin & M. Reza Ghadiri A self-replicating peptide Nature 382, 525 - 528 (1996).
  10. Stuart A. Kauffman (1996) Self-replication: Even peptides do it. Nature 382 August 8, 1996.
  11. David H. Lee, Kay Severin, Yohei Yokobayashi & M. Reza Ghadiri Emergence of symbiosis in peptide self-replication through a hypercyclic network. Nature 390, 591-594 (1997) Letters to Nature.
  12. David H. Lee, Kay Severin, Yohei Yokobayashi & M. Reza Ghadiri : Correction: Emergence of symbiosis in peptide self-replication through a hypercyclic network. Nature 394, 101 (1998) Letters to Nature.
  13. Lou Allamandola(1998) Stuff of life, New Scientist, 12 Sept 1998, pp30-35.
  14. Biliang Zhang, Tom Cech (1998) Molecules of ancient life are born again, New Scientist, 17 Oct 1998, p10.
  15. G. von Kiedrowski : A Self-Replicating Hexadeoxynucleotide Angew. Chem. 98, 932-934 (1986).
  16. Shao Yao, Indraneel Ghosh, Reena Zutshi & Jean Chmielewski: Selective amplification by auto- and cross-catalysis in a replicating peptide system. Nature 396, 447 - 450 (1998). 3 December 1998.
  17. The estimates of the total number of human genes range from 30,000 - 120,000 with a mean of 62,598 (see page of Ewan Birney). But why not start with the known number of cell types 256 and predict the unknown number of human genes as 2562 = 65,536 ? In 2003 we know the answer! In Feb 2001 the estimates moved to 30,000 - 40,000.
  18. John Maynbard Smith & Eörs Szathmáry(1999) The Origins of Life, p7. (review on this site)
  19. See also about the origin of life: The Principles of Life by Tibor Gánti (review on this site).
  20. Tracey A. Lincoln and Gerald F. Joyce (2009) 'Self-Sustained Replication of an RNA Enzyme', Science, 27 February 2009. ("This has now been achieved in a cross-catalytic system involving two RNA enzymes that catalyze each other's synthesis from a total of four component substrates.").

Further Reading

  • Kauffman's home page
  • Kauffman's home page at the BIOS group
  • Stuart Kauffman (2006) "Intelligent Design, Science or not?" in: Intelligent Thought: Science Versus The Intelligent Design Movement edited by John Brockman (info). This is an important and insightful essay.
  • Investigations: There is new experimental evidence for self-replicating peptides, DNA and RNA molecules in his new book "Investigations".
  • Essay by Stuart Kauffman about his new book Investigations at bn.com
  • Stuart Kauffman (2008) Reinventing the Sacred: A New View of Science, Reason, and Religion, Basic Books, 320 pages. Reviewed in Science, 20 June 2008 by Denis Noble. Also reviewed in: Scientific American by Michael Shermer. See also online: Stuart A. Kauffman BEYOND REDUCTIONISM. Reinventing The Sacred, Edge. The Third Culture.
  • Announcement of Investigations in "The Editors Recommend" (Scientific American, January 2001).
  • New Scientist interviews Stuart Kauffman about Investigations, New Scientist 17 March 2001, p46-49.
  • Let there be life by Phil Cohen, New Scientist, 1997
  • Forget In Vitro - Now It's "In Silico", Interview with Stuart Kauffman by Ken Howard, Scientific American, June 5 2000, pp50-51.
  • Julie Wakefield: Complexity's Business Model, Scientific American, January 2001 (includes Kauffman's Bios Group).
  • Elizabeth Pennisi: 'Systems Biology: Tracing Life's Circuitry. Science, Volume 302, Number 5651, Issue of 5 Dec 2003, pp. 1646-1649.
    A new movement aims to integrate biology, mathematics, and engineering; even if its objective is hard to define, it is all the rage in the academic world.
  • Review of The Principles of Life by Tibor Gánti (on this site).
  • A Third Window: Natural Life beyond Newton and Darwin (Paperback) by Robert W. Ulanowicz (Author), Stuart A. Kauffman (Foreword) (2009): "Chance and disarray in natural processes are shown to be necessary conditions for real change. Randomness is shown to contribute richness and autonomy to the natural world."

Reviews: 		 


guestbook (moderated) home: www.wasdarwinwrong.com http://home.planet.nl/~gkorthof/kortho32.htm  
Copyright © 1998 G.Korthof . First published: 20 Oct 1998 updated: 5 Mar 2009 Notes/F.R. 27 May 2009