Self-organization
Some scientists have suggested that something more than chance and natural selection is involved in the linking of amino acids to form proteins. They propose that certain chemical systems have self-organizing properties or tendencies. Steinman and Cole (1967) suggested that one amino acid may be attracted to another amino acid more than it is attracted to others. There is experimental evidence that this is true. There is some differential attraction among amino acids. Steinman and Cole claimed that the ordering of amino acids they observed in their experiments matched the ordering of amino acids in ten actual proteins. But when Bradley and his coworkers (Kok et al. 1988) compared the sequences reported by Steinman and Cole to a larger sample of sequences from 250 actual proteins, they found these 250 sequences “correlate much better with random statistical probabilities than with the experimentally measured dipeptide bond frequencies of Steinman and Cole” (Bradley 1998, p. 43). Also, if the properties of the twenty biological amino acids strongly determined the bonding of protein sequences we would expect only a few kinds of proteins to form, whereas we observe that thousands form (Bradley 1998, p. 43).
Another kind of self-organization happens when disordered molecules of a substance form crystals. This is technically called “spontaneous ordering near equilibrium phase changes.” The formation of crystals is fairly easy to explain. For example, when the temperature of water is lowered below the melting point, the tendency of water molecules to interact in a disordered way is overcome, and they link together in an ordered fashion. In this phase transition, the water molecules tend toward a state of equilibrium, moving to the lowest potential energy, giving up energy in the process. Imagine that there is a large depression in the middle of a billiards table. If you tilt the table here and there, the wandering balls will naturally wind up in the depression, touching each other and motionless. In the process energy is lost i.e. the process is exothermic. But the formation of complex biological molecules (biopolymers) is different. It is an endothermic process, meaning heat is added, and it takes place far from thermal equilibrium. The polymers are at a higher energy potential than their individual components. It is as if the pool table has a hump in the middle, rather than a depression. It is a lot more difficult to imagine all the balls winding up together on top of the hump simply as a result of random movement, than it is to imagine them winding up in the depression in a state of thermal equilibrium. It would take some energy to get the balls up on to the hump and keep them there. Bradley (1998, p. 42) says, “All living systems live energetically well above equilibrium and require a continuous flow of energy to stay there . . . Equilibrium is associated with death in the biosphere, making any explanationof the origin of life that is based on equilibrium thermodynamics clearly incorrect. . . . phase changes such as water freezing into ice cubes or snowflakes is irrelevant to the processes necessary to generate biological information.”
The kind of order found in crystals is repetition of simple patterns, whereas the kind of order found in living things is highly complex and nonrepetitive. The order found in the biochemical components of the bodies of living things is not only highly complex, but very specific. This specified complexity has a high information content, which allows the biochemical components to perform specific functions that contribute to the survival of the organism. Compare the letter sequences ABABAB AB, RXZPRK LDMW, and THE BIG RED HOUSE. The first sequence is ordered, but it is not complex and therefore is not informative. Crystals are like this. The second sequence is complex, but it is also not informative. But the third sequence is both complex and informative. The sequence of letters encodes information that allows the sentence to perform a specific communication function. This property can be called “specified complexity.” Biological complexity of the kind we are talking about in proteins and other molecules in cells is specified complexity— it is complexity that specifies a function (like protein coding ability of DNA). Such patterns of complexity are thus different from the simple repetitive patterns that arise in the crystallization process (Meyer 1998, p. 134).
Prigogine proposed that self-reproducing organisms could arise from reacting chemicals brought together in the convection currents of thermal baths, far from thermal equilibrium. This is somewhat different from the crystal formation process, which involves phase transitions at or near thermal equilibrium. Bradley (1998, p. 42) nevertheless concludes that although the ordered behavior of the chemicals in Prigogine’s systems is more complex than that observed when the systems are at thermal equilibrium, the order is still “more the type of order that we see in crystals, with little resemblance to the type of complexity that is seen in biopolymers.” And whatever ordering is observed can be attributed to the complex design of the experimental apparatus. Meyer (1998, p. 136), citing the work of Walton (1977), says, “even the self-organization produced in Prigogine-like convection currents does not exceed the organization or information represented by the experimental apparatus used to create the currents.”
Manfred Eigen has proposed that groups of interacting chemicals called “hypercycles” could be a step toward self-reproducing organisms (Eigen and Schuster 1977, 1978a, 1978b). But John Maynard-Smith (1979) and Freeman Dyson (1985) have exposed some flaws in this proposal. “They show, first,” says Meyer (1998, p. 136), “that Eigen’s hypercycles presuppose a large initial contribution in the form of a long RNA molecule and some forty specific proteins. More significantly, they show that because hypercycles lack an error-free mechanism of self-replication, they become susceptible to various error catastrophes that ultimately diminish, not increase, the information content of the system over time.”
Stuart Kauffman of the Sante Fe Institute has tried another approach to complexity and self-organization. He defines “life” as a closed network of catalyzed chemical reactions that reproduce each molecule in the network. No single molecule is engaged in self-replication. But he asserts that if you have a system of at least a million proteinlike molecules, the odds are that each one will catalzye the formation of another molecule in the system. Therefore the system as a whole replicates. When the system reaches a certain state, it supposedly undergoes a phase transition, introducing a new level of complexity for the whole system. But Kauffman’s concept is based purely on computer models with little relevance to real life systems of reacting chemicals (Bradley 1998, p. 44).
First of all, Kaufmann’s estimate of a million molecules is too low for each kind of molecule to catalyze the formation of another kind of molecule in the system. But even if a million kinds of molecules is enough, the odds that a particular catalyzing molecule will be near the correct chemical ingredients needed to produce another molecule are remote (Bradley 1998, p. 45).
Futhermore, Kaufmann’s computer models do not adequately take into account the exothermic nature of the formation of biopolymers— the reactions require energy from the system and would quickly deplete it, leaving the system “dead.” Kaufmann proposes that energy-producing reactions in the system could compensate for the energy consumed in the formation of biopolymers. But Bradley (1998, p. 45) points out that these reactions will also require that certain molecules be in the right places at the right times, in order to participate in the reactions. How all this is supposed to happen is not satisfactorily explained in Kauffman’s models. Bradley (1998, p. 45) adds: “Dehydration and condensation onto substrates, his other two possible solutions to the thermodynamic problems, also further complicate the logistics of allowing all of these 1,000,000 molecules to be organized into a system in which all catalysts are rightly positioned relative to reactants to provide their catalytic function.” In other words, Kauffman’s system does not realistically account for getting all the molecular elements arranged in the proper places for all the needed catalytic and energy-producing reactions to take place. In a computer this may not matter, but in real life it does.