The Failures of Mathematical Anti-Evolutionism
Download 0.99 Mb. Pdf ko'rish
|
The Failures of Mathematical Anti-Evolutionism (Jason Rosenhouse) (z-lib.org)
(Kimura 1961, 127)
Statements of this sort are ubiquitous in the literature of evolutionary biology, but notice that it is specifically genetic information that is intended here. Standard evolutionary mechanisms have the ability both to increase the information storage capacity of the genome through gene duplication, and to refine, via mutation and natural selection, the information stored there, resulting in better-adapted organisms. This is what is meant when scientists attribute to evo- lution the ability to create new information. If you now want to play gotcha, and argue that evolution did not really create information, but only transformed preexisting information in the environment, then you are welcome to do so. However, it is no great accomplishment to make this observation. If you are just saying that nature has to be a certain way for evolution to work then you can just assert it as obvious without further argument. You do not need to write lengthy books to defend this claim, or to deploy difficult mathematical theorems in support of it. However, Dembski wants to go further. In his telling, genomes do not merely contain information in some general sense, but instead contain complex, specified information (CSI), which is supposed to be an indicator of intelligent design. He sees his NFL-inspired argument as a rebuke to scientists who claim that evolution can produce both complexity and specification. To the extent that evolution can do this at all, he argues, it is only because the environment must have contained CSI to begin with. He refers to this as “the displace- ment problem,” his idea being that scientists have explained CSI in 206 6 information and combinatorial search genomes only by helping themselves to a source of preexisting CSI. He likens this to filling one hole by digging another. We still have to explain the environmental CSI, he argues, and for this purpose we must invoke the action of an intelligent designer. But Dembski has a lot more work to do if this argument is to be taken seriously. Recall that within his formalism, demonstrating the presence of CSI requires two things: We must carry out a relevant probability calculation to establish complexity, and we must also describe a proper specification in the precise technical sense we saw in Section 5.7. We have already seen that Dembski has no way of carrying out either of these steps for biological adaptations, and he will have no more luck here. Calculating the probability that the fundamental constants would have just the values they do entails knowing the appropriate probability distribution, but we have no useful information for deciding on what distribution that is. Nor do we have any relevant background knowledge that would help us jump through the mathematical hoops required by Dembski’s notion of specification. Until Dembski provides these details, he has no basis at all for claiming that the environment contains CSI in his idiosyncratic sense. We should note that in Dembski’s trichotomy of regularity, chance, or design, both regularity and chance are viable options for explaining the origin of the constants. Modern cosmology can tell us almost nothing about how the constants came to be what they are. It could well be that some sort of natural law latent in the origin of the universe determined that the constants had to take on the values we observe, in which case regularity would be the explanation. To illustrate my point, suppose I told a friend that I tossed a coin 100 times and that it came up heads every time. My friend might reply, “That’s amazing! 100 consecutive heads is incredibly unlikely.” But now I point out that, actually, I used a two-headed coin. What at first seemed incredibly unlikely is seen to be inevitable once all of the relevant information is at hand. I am suggesting that the constants of the universe might have the values they do simply because some 6.9artificial life 207 unknown physical principle requires that they do. There is nothing in current cosmology to even suggest, let alone imply, otherwise. Alternatively, it could be that the constants really might have set themselves to a large number of different values, but that ours is just one universe in a vast multiverse, each with its own set of constants, and in this case chance would be the explanation. This scenario can be analogized to winning a lottery. If we only buy one lottery ticket then we are unlikely to win, and this is analogous to finding just the right set of life-supporting constants in a single universe. But if we buy billions of lottery tickets then it is near-certain that at least one will win. This is analogous to finding one life-supporting universe in an enormous collection of such universes. Dembski needs to eliminate both of these possibilities for his argument to have any force, but he has no plausible way of doing so. Summing up, the point is that when stripped of all of the irrel- evant mathematical formalism, there is nothing more to Dembski’s argument beyond wondering why the universe has just the properties it does, and then to assert, based on nothing, that design is the only possibility. The only role for the NFL theorems is to justify the claim that the environment must be a certain way for Darwinian mech- anisms to be effective, but that claim is obvious without invoking difficult mathematical formalism. It is an interesting question to ask why the universe is as it is, but it is hardly a problem in biology’s domain. Thus, even taking Dembski’s argument at face value there is nothing here to which evolutionary biologists need to pay attention. 6.9artificial life The final ID argument we will consider in this chapter unites their thinking about information theory with their rhetoric about combi- natorial search. We have one last piece of groundwork to lay before we come to that. 208 6 information and combinatorial search As compelling as the circumstantial evidence for evolution is, it would be better to have direct experimental confirmation. Sadly, that is impossible. We have only the one run of evolution on this planet to study, and most of the really cool stuff happened long ago. If we someday find life on another planet, then we will be able to increase the sample size to two, but that seems unlikely to happen any time soon. Experimental evolution on microorganisms is possible since they can be cultivated in large populations and reproduce very quickly. Small animals like fruit flies also lend themselves to such experimentation for the same reason, as do plants. Evolutionary experiments on larger animals have been undertaken, but they are limited by a host of obvious practical problems. In short, there has been a lot of illuminating research in the realm of experimental evolution, but by its nature it has little to tell us about the big questions of natural history. For example, you are not going to witness the evolution of a large-scale organ system in the course of these small-scale experiments. However, Darwinian evolution occurs in any environment in which there is heritable variation, replication, and competition for resources. This suggests the possibility of creating an artificial envi- ronment that features those three characteristics. Studying evolution in such an environment could illuminate general principles that arise in any evolutionary environment. Based on such experiments, we might hope to say, “Here’s what happened in our synthetic environ- ment, this environment reproduces relevant aspects of the natural environment, and therefore it is plausible to think something similar has happened in natural history.” In particular, we can use computers to create this environ- ment. Our organisms will be computer programs striving to replicate themselves and competing for memory in the computer. We can introduce something equivalent to random mutations, so that as programs replicate themselves there is some possibility of alterations to their code. And we can modify the environment to set up selection 6.9artificial life 209 pressures in favor of certain abilities the programs might be able to develop. Experiments of this sort have been carried out since the late 1980s, and this area of research is known as “artificial life.” It is commonplace, even in the professional literature, to refer to artificial life experiments as “simulations of evolution.” This is convenient terminology for distinguishing these experiments from what happens in nature, and we will use it ourselves, but we should note that it is also slightly misleading. Experiments of this sort are not so much simulations of evolution as they are instances of it. In observing such an experiment you are watching actual evolution take place, albeit in an environment in which the researchers control all of the variables. An early example of the genre was an experiment carried out by Thomas Ray using a platform he dubbed “Tierra.” He started with an 80-line computer program, written in a computer language of his own devising, capable of replicating itself. The only selection pressure was towards efficient replication. Yet even in this incredibly simple environment, astonishing complexity quickly evolved. In a commen- tary on this experiment, biologist John Maynard Smith described what occurred (note that in the quote, a “satellite virus” is a virus that is completely dependent on its host’s machinery in order to replicate): The evolutionary behaviour of the system is surprisingly rich. The first important variants to arise are ‘parasites’, 45 instructions long, which cannot replicate on their own, but use the copying procedure of a neighbour: they are an exact analogue of satellite viruses. Once parasites become common, ‘hosts’ may evolve immunity; then new types of parasite evolve that are able to attack the new hosts. Evolutionary arms races of this kind have often been postulated and occasionally observed. Next there are ‘hyperparasites’ which, by an ingenious trick, persuade the parasites to replicate them. Download 0.99 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling