Richard Dawkins' Evolution by Computer: Randomness Rescued?


One method of argument is the snow job and I am not saying evolutionists always use it deliberately, but if something sounds complicated and deep, it is easy for people to think, "Wow, I guess he really solved that problem!"


When creationists present the argument that the improbable structures needed for living organisms could not be created by random events, evolutionists often refer them to a computer model designed by Dr. Richard Dawkins that makes an unlikely task much more likely by breaking it down into steps in which successive approximations are selected.  For example, this is the tack taken by Dr. Edward Max in his article  “The Evolution of Improved Fitness By Random Mutation Plus Selection.”  (see

Such an argument  has apparently seemed convincing enough to some creationists that there is now a bit of reluctance to use the probability argument.  Let us take a closer look and see what has really been proved.  The following is a section from an on line debate I had with Dr. Max. (see )


The task Dr. Dawkins used is the generation of the Shakespearean phrase, "Methinks it is like a weasel," by successive random variations. He started with the necessary characters and spaces and randomly programmed the computer to generate a random sequence from a "soup" of the component letters and spaces used in that sentence. The letters and spaces could be used more than once but the length was made to be precisely 28 objects. He got "WSLMNLT DTJBKWIRZRESLMQCO P." This was to be the starting point for computerized simulation of natural selection.


In a sense it was like abiogenesis, except Dr. Dawkins apparently feels he does not have to deal with the origin of the first living creature. If he did, he would have to deal with the problems of "d" ("right handed") and "l" ('left handed") amino acids being the norm in nature but only the "l" forms being used in life. He would also have to explain the fact that they do not come together in long chains outside a living cell or biochemist's test tube. If Dawkins' first series were really like abiogenesis, it would have to be a sentence that made sense because it would have to be a sort of primitive version of Shakespeare.  Do Dr. Max and Dr. Dawkins feel that this is Australopithecine iambic pentameter? (Come to think of it, I recall a line like this from an early "I Love Lucy" episode... no, wrong "Lucy.")


Subsequently, the computer was reprogrammed for the next step. Let me use Dr. Max's words, "Then, following Dawkins revised program, the computer made multiple copies (progeny) of this sequence, while introducing random 'errors' (mutations) into the copies. The computer examined all the mutated progeny and selected the one that had the most similarity (however slight) to the line from Hamlet."


After 30 generations of this procedure, Dr. Max reports, the sentence had "evolved" from the original to "MDLDMNLS ITJISWHRQREZ MECS P." and by the forty-third generation it had arrived at "METHINKS IT IS LIKE A WEASEL." Firstly, since the sample sequences do not have the same numbers of each of the letters and spaces (and assuming that they are actual data from the experiment) I suppose that the program allowed substitutions of any of the characters and was not just a shuffling of the correct ones. That, of course, made it harder to arrive at the desired answer, but the fact remains that several assumptions have to be made to accept this simulation as equivalent to evolution.


Since Dr. Dawkins is a wit and also British, there may have been a reason for his choice of quote. For, indeed, there is a bit of the weasel operating here. First of all, as Dr. Max admits, this presupposes a reproducing system and fits nicely with his desire to just skip the whole question of the origin of life. But secondly, because the selection was narrowed 43 times by selecting the "closest" sequence to the goal, it presumes that there is some equivalent mechanism involved in nature. "Of course," he says, "It is natural selection!"


Does close count in horseshoes, hand grenades AND mutations? Only if the closeness does something helpful. Now, operating in reverse, if you have a working protein and make some substitutions or deletions of amino acids, it turns out that there are quite a few of those variations that do not harm the function of the molecule. That is equivalent to spelling variations that may be permissible or spelling errors that are easily understood. Some spelling mutations, however, cause grave problems, like "You must not go out," changed to "You must now go out." Protein alterations are similar.


At the risk of having Dr. Max claim that I know nothing about proteins or Dr. Dawkins' experiment, I will push the analogy further. Suppose you want to write the Encyclopedia Britannica by putting millions of monkeys on laptops. Will you find that after a while, one of the monkeys begins to make a little sense so that you are able to sell a few copies (perhaps to college undergraduates), then fund the whole zoo to work on that version until it makes even more sense? For the program to reflect reality, there must be some clear advantage at each stop along the way, followed by the near elimination of all the other competing configurations until further mutations produce another with even more advantage.


This may sound exactly like evolution, and one who already believes it will cry "Eureka!" But familiarity of a concept does not make it a fact. This may come as a shock to some of you, but things do not necessarily "go better with Coke." First of all, without a system that reproduces, there is no natural selection at all. It is easy to presume that there was "an early stage of proto-life" as Dr. Max does, and even thinks that it allowed more of the larger and more dramatic mutations because it was less accurate than replication is today. Reproduction is a very complicated process (despite the fact that some people supposedly do it "by accident") and only the discovery of the DNA molecule made it understandable. Yet the double helix DNA, containing two strands, each of which is essentially the chemical "mirror image" of the other, must also be regulated and helped by a complex system in order to open up in the proper way and attract the nucleotides that compliment both the "sense" strand and the "anti-sense" strand and then separate and move into two daughter cells.


If Dr. Max and those who agree with him can even imagine another system, much less a "simple" system for duplication of a complex molecule, let us hear about it. You see, the very impossibility of doing it is precisely evidence that it DID NOT happen by natural means. Dr. Max says he might just leave that up to God. But he still thinks Creationists are sloppy thinkers.


He thinks it is obvious that DNA varies, causing proteins to vary until they are able to do something significant for the cell. Then they become desirable so that and "everybody" has to have them. After a while another permutation becomes even more effective at something and becomes the new standard. Now, I think this sounds a lot like the evolution of adolescent fashion, and it is highly debatable whether THAT represents improvement. But as to coming upon solutions to metabolic problems of an organism by trial and error, it really stretches the imagination. Yet, in his mind, it is the Creationists who are out to lunch.


Surviving organisms, doing all this variation, have to survive. That means that the variation must not interfere with other things the cell needs to do. And there are a lot of tasks the cell needs to do to survive. Notice that I say, "cell." Some say that there was some simpler "pre-cell" life that did not need all this complexity, but there are certain tasks that must be done for a "thing" to qualify as life. Viruses are not alive because they need to hijack the machinery of a living cell to reproduce themselves. They are more like fragments of cells that turn on cellular activity when they find the proper spot from which to function.


You can have a simple car that would do the minimum required of a car nowadays, like start, run, steer and stop. You could do without some common features such as sound system (unless the user is under 30), windshield, roof, inflatable tires, or even seats, but eventually you get to the point where you cannot strip down any further and still get from here to there. And in strip down cells, the most basic model is still incredibly complex. Darwin and Huxley did not know that "protoplasm" was quite a bit more complicated than Jell-O, but present day molecular biologists do not have any excuse. And wishful thinking is not a valid answer -- "We hypothesize a simpler, more primitive self replicating system." That, I say to all you lay people who are following this line of argument, is indeed a snow job by the experts, maybe even science fiction.


But we already HAVE the living system, Dr. Max reminds us, and it just improves itself. To use the car illustration, the engine starts to mutate from a one cylinder to four or six or eight (and then, of course, back down again as the environmental gas supply is depleted.) Are there going to be intermediate stages that function better? If you have two pistons but only one cylinder, will it run better? No, it won't run at all. It is hard to even imagine a scenario where slow changes, or even fast changes - for those of you who think that a change in a regulatory gene can do some amazing things - can get from here to there by way of changes that all work and work better than their predecessors. Try as an exercise changing a reciprocating engine into a jet engine. Without other accompanying changes in all sorts of systems, even the addition of a working new feature will not help, and might tear the whole thing apart, like putting a jet engine on the Wright Brothers' Plane.


But, I think I can hear Dr. Max protesting. The example is faulty and typical of Creationists, oversimplifying things so that the laity think they understand. The biological example consists of DNA doing the alteration and because of this, producing alterations in parts of the cell. It sounds absurd to think of autos mutating and makes untrained people think that the experts are foolish.


Let me therefore use another example to help us all, including Dr. Max, understand my objection.  Think of the cell as a computer system, perhaps even including a monitor, printer, modem, scanner and even connections to your furnace thermostat, clock radio, coffee pot and a variety of other task-oriented mechanisms. For, indeed, the cell does not just compute but does various forms of work and interacts in certain ways with the outside world. But the computer also contains a lot of things that seem extraneous. There may indeed be worthless programs and files, downloaded by accident or by someone without sales resistance, clogging up your hard drive. (Creationists would regard this as a perfect illustration of deterioration and having nothing to do with improvement.)


Suppose you are unhappy with the performance of your computer and want to improve it. So you start to make random changes, deletions, alterations and substitutions in the computer code. Are you likely to help things along? I doubt it. In fact, the most likely results are either no change, loss of a function or complete lock up. As one who used to try (on my old "coal fired" prehistoric DOS Computer) -- with the manual in one hand -- to add or subtract a line in "config.sys" or "autoexec.bat," let me say that being close but wrong was often no help whatsoever in preventing a major disaster.


So, in summary, Dr. Dawkins example of computer generated evolution will only simulate evolution if each of the steps along the way to the statement "METHINKS IT IS LIKE A WEASEL" were a meaningful sentence.  But the steps were only intermediates that were selected because they were closer to the goal and evolution does not know about goals.  It only selects things that are better in the present.  And in the final analysis, “WSLMNLT DTJBKWIRZRESLMQCO P” actually no better than “WSLMNLT DTJBKWIRZRESLMQCO P” and neither makes sense.


Interestingly, when confronted by this argument, Dr. Max denied that Dawkins' computer evolution model, starting with random letters and ending up with "METHINKS IT IS LIKE A WEASEL" proves or supports evolution and stated that he never said otherwise. Rather, he only claims it shows that unlikely results can be more easily produced by multiple successive steps of approximation than by a single chance event. 


He can claim the limited objective of proving that narrow point, but cannot apply it to the place he really needs it – evolution. 


So, should creationists be shy to use the probability argument?  Of course not!  The counter-argument is just a snow job, and right now is a good time for a major thaw.


                                                                                                Ross Olson MD