When I got interested in evolution, one of the first books I read was The Blind Watchmaker, by Richard Dawkins. I had never heard of Dawkins before reading that book. I read it simply because I happened to notice it a the public library and thought it had a cool cover.
The book's third chapter is called, “Accumulating Small Change.” In it, Dawkins explains how the prolonged action of natural selection can craft complex structures. He introduces his famous, “Methinks it is like a weasel” experiment. More precisely, he does two experiments. In the first, random strings of characters are generated until the weasel phrase is obtained. In the second he starts with a few randomly generated strings. Then the few strings that happen to have any resemblance to the target phrase, even if it is just having one letter in the right place, are used as the progenitors for the next generation of phrases, in a process modeled after genes mutating in the offspring of the current generation. This process continues until the target phrase is reached.
Of course, the first experiment fails completely to reach the target phrase while the second experiment reaches it very quickly. Dawkins's point was that many people naively think evolution by natural selection proceeds as in the first experiment, while in reality it is much closer to the second.
When I first read the book I was disappointed. Dawkins devotes an awful lot of space to establishing what seemed to me an obvious point. Yes, of course natural selection alters the probabilities of arriving at complex structures. Of course natural selection could, in principle, lead to the accumulation of small change. Why was he belaboring this?
It was when I started reading creationist replies to Dawkins that I came to understand why Dawkins had gone on at such length. No matter how slowly and clearly you explain this simple point, even the classier creationists try to argue against it.
I was thinking about that as the result of the recent debate between Stephen Meyer and Lawrence Krauss about God and science and whatnot. I have not had a chance to watch the video of the debate, but it has certainly generated a lot of internet chatter. I gather that at some point Meyer, representing the creationists, whipped out probability arguments to make his case. Having read his books, I am aware that he is very taken with such things.
Not having seen the debate, this post is not a reply to Meyer. But I do notice that Richard Dawkins has left a comment over at Jerry Coyne's website about the debate. He writes:
When will these people understand that calculating how many gazillions of ways you can permute things at random is irrelevant. It's irrelevant, as Lawrence said, because natural selection is a NONRANDOM process. You'd think they’d realise that if it were THAT easy to disprove evolution no scientist would take evolution seriously. Do they really think we are so very stupid? Or are they cynically playing to the gallery, dazzling the naive audience with big numbers like 10^77, while knowing full well they are irrelevant?
Another commenter at Jerry's site then wrote:
I think I can play this game. How's this for dazzling? I walk a mile from my train station to my office every day. If I just randomly took enough steps to walk a mile, I could end up anywhere in a 3.14 square mile area. My cubicle is only about 20 square feet. This leaves about a one in 4.4 million chance of finding it. I’ve done this 1000 straight days which equates to a 1:10^6641 chance of this having randomly happened.
The obvious conclusion is that my commute isn’t random. Why is it so hard for these people to come to the same conclusion about evolution, especially when it is repeated to them ad nauseam?
Indeed. But now we come to the point of this post. You see, Douglas Axe of the Discovery Institute then stopped by to defend the creationist view of this:
Equally obvious is that your commute isn't random because you know where you're going. When I read The Blind Watchmaker as a student (and enjoyed it very much, I might add) I was under the distinct impression that “blind” meant NOT knowing where one is going.
So, let's look at your analogy again. You receive benefit in the form of a paycheck for showing up at your cubicle every work day. If instead you were to meander blindly from the train station (meaning, without using any knowledge of where you’re supposed to go) do you really think your employer would give you 0.1% of your pay if you happened to end up 0.1% closer to your cubicle than your starting point? And then 0.2% if you happened to make it a little closer the next day?
Of course you don't. You get paid after making it the whole way and doing your work.
That's the glaring problem with natural selection. It's huge weakness has nothing to do with randomness and everything to do with the fact that it only rewards good work AFTER that work has been done. Consequently it has absolutely nothing to do with HOW that work gets done in the first place.
Speaking of ad nauseam repetition, this obvious shortcoming of natural selection has been repeated over and over for well over a century. As it was put in a quote offered by Hugo De Vries in 1904: “Natural selection may explain the survival of the fittest, but it cannot explain the arrival of the fittest.”
Over at the Discovery Institute, David Klinghoffer thought this was a clever point.
But Axe is just being silly. Since he mentions Dawkins's book, let us start there. In both of Dawkins's experiments, the computer knew the target phrase. But it was only in one of them that the target phrase was found.
Plainly, then, it is not knowing the target phrase that guarantees success. Rather, it was the presence of a consistent selection mechanism, coupled with copious variation, that led to success in the second experiment.
The question, then, is whether evolution by natural selection mimics the relevant parts of Dawkins's experiment. Clearly, it does. Large populations of organisms produce enormous quantities of new genetic variations in each generation, and natural selection then consistently selects for variants that increase the probability of survival. Given these two facts, the conclusion of Dawkins's experiment stands. Such a process can bridge vast gulfs of improbability.
Axe's point therefore has no force. He says that natural selection has “has absolutely nothing to do with HOW that work gets done in the first place.” Whether that is true depends a lot on what you mean by “that work.” Presumably the work in question is crafting a complex structure gradually over long stretches of time. But I'd say that natural selection has a mighty big role to play in how that work gets done.
Axe writes as though mutations are one thing and selection is something different. Natural selection, he says, has nothing to with it, because it can only select from the variations put before it. In his comment, “the work” is done before selection even arrives on the scene. But this is obviously not the right way of looking at it. There is only one evolutionary process under discussion here: It is the one where random genetic variations are passed through the sieve of natural selection. It is both together that can produce complex, directional change. It makes no sense to discuss either one as though it is separate from the other.
If you want to argue that natural selection cannot craft complex structures, then you cannot do it from your armchair. Anti-evolutionists are constantly trying to make their point on the cheap. They gesture futilely toward the second law of thermodynamics, or they blather about “irreducible complexity&rdquo, or they hold forth on how natural selection is a meaningless tautology, or they carry out back of the envelope probability calculations. These arguments are nonsense. Their ubiquity in anti-evolutionist writing goes a long way to explaining the general contempt scientists feel for it.
There is no theoretical reason why natural selection acting on small variations cannot, over time, produce complex structures. You will not get by with “in principle” arguments against it. To know if it is plausible in specific cases, like eyes or flagellae or immune systems, you will have to do the hard work of studying those systems. Are they pristine creations from nothing, or are they cobbled together from preexisting parts? Are they engineering marvels, or are they Rube Goldberg machines that show the “senseless signs of history,” as Stephen Jay Gould put it? In each case the answer is the latter. Far from being a challenge to evolution, those complex structures creationists endlessly go on about provide strong confirmation. They have just the structure they would need to have for evolution by natural selection to be a viable explanation.
There's plenty more to be said about this, of course. In the time I've been writing this post, I've noticed that Stephen Meyer has written his own reply, and I'd like to watch the debate before commenting further. But there is no time for that now. I have tickets to the midnight show of Batman vs. Superman...
That's all true, but there is still an argument to be made against Dawkins' weasel. His computer has, at the start, a complete specification of the end it is working towards, and therefore its process is not a good analogy for natural selection. No specification of a weasel exists in nature until the process of selection has produced an actual weasel.
Dave Thomas has an article entitled "War of the Weasels" in Skeptical Inquirer Volume 34.3, May / June 2010 which uses a genetic algorithm to solve a math problem called Steiner Problems, and he specifically addresses this failure of Dawkins' analogy. In Thomas' paper, the problem has no target given - just a fitness measure - and if there were, he claims, then a creationist should be able to produce the solution from that target easily. Spoiler alert: they couldn't.
The genetic algorithm found solutions very near the optimum, and a few "MacGyver" solutions - weird hacks that were close to optimum. A great analogy to, for example, the recurrent laryngeal nerve.
Pointless. Creationists and religious people by definition lack basic reason. You can't reason sight into a blind person. Creating gods in our image has brought nothing but pain, suffering and death to humanity. Facts are meaningless to such simple minds.
I am disappointed that anyone would stoop to browbeating tactics as name calling, blame shifting, and steriotyping.
The search for an true and acceptable explanation for the origin of life meaning for all people - simple and educated.
I am a specialized critical care registered nurse by trade, creationist by education, and a Christian by religion.
I understand biology, anatomy, physiology, pathphysiology, pharmicokinetics, etc than most people in the world and hold that natural selection and evolution are not the best explanation for the origin of life. I am not blind, ignorant, angry, or unreasonable. Making sweeping statements such like what is above, isn't helpful. It only shows your bias, disrespect, and disregard. None of that will add credibility or an audience.
Let’s look at Axe’s contribution here: Whilst neither the “Methinks it is like a weasel” analogy nor the “commute to work analogy” is quite perfect in its illustration of evolution by natural selection (ENS), the objections that Axe raises are clearly not applicable. He seems to think that by specifying in advance the desired outcome – namely getting to work or happening upon a Shakespearian quote – then the analogy will not hold for ENS where the outcome is, by definition, not specified in advance. But this is to miss the essential point. In ENS the “desired outcome” (which needn’t be defined, or even stated, in advance in any specific terms) will be determined after the event by whether or not the newly changed organism is or is not more fit to survive in the environment in which it finds itself. What the details of that environment happen to be are irrelevant to the argument – if the environment happens to be high in CO2 then adaptations that deal with that will be selected. If the environment is low in CO2 and also low in oxygen, then adaptations that cope with hypoxia will be selected. The point is, it doesn’t matter which 6 word Shakespearian quote we happen to be aiming for in the first place (and whether or not we specify this from the start, or even whether we know or not which quote it is we’re actually looking for), whatever is selected either by the computer (or by ENS, or by the commuter trying to get to work, or by whatever), as long as it is not randomly selected but is instead subject to some sort of selective pressure, we can generate adaptation by trial and error. Indeed if Dawkins, as Axe would seem to want, were to be somewhat less specific and merely demand that his programme come up with “any consecutive phrase of 6 words that appears somewhere in the manuscripts of Shakespeare”, then we would undoubtedly arrive at a resolution rather more quickly. In other words, in his programme, Dawkins is perfectly entitled to decide whatever environmental pressure he wishes to propose – all he need do thereafter is to demonstrate that in whatever case he’s chosen, then selective processes working on random variation will come up with the goods – whatever the end point may be. This is not only necessary to make his point, it is also sufficient.
IMHO it is inconceivable that Axe does not understand the fallacy in his own argument. He is a well trained scientist who has published in the peer reviewed literature, and he is undoubtedly both intelligent and well educated. (See his entry in “Encyclopedia of American Loons”):-
Why would such a person insist on being so obtuse in the face of a cogently argued and well presented position, backed up by mountains of real world evidence? Perhaps there is a serious neuropsychological thinking disorder embedded deep within the human genome, which finds regular expression in a significant percentage of our species? - a disorder which (in at least some cases) no amount of education or rational argument can overcome. Consider, for example, this curious little exchange which appeared as a “post of the month” over at Talk Origins sometime in 2009:-
Furthermore, assuming that such a cognitive deficit does indeed exist in our chemistry, then why has it not been selected out? Curious that such a gaping hole in our fitness should have persisted for so long. Perhaps the need to establish “truth” with respect to the nature of the world and universe within which we live, is not all that important. A horrible notion – but I guess that’s another discussion.
"I understand biology..." You may understand it better than most but clearly not enough. "... creationist by education, and a Christian by religion." I think I see your problem.
Another thought - instead of considering Axe's argument as applied here, let's take a very good real world example in actual biology, using real living thngs - Lenski's experiment. Apply Axe to E.Coli adapting to a citrate rich environment, and it immediately becomes apparent just how hollow his line of reasoning actually is.
David Evans wrote:
"His computer has, at the start, a complete specification of the end it is working towards"
This specification is just a success measure and it's an arbitrary one. We can swap it out for "produces more children" and make it a closer match to natural selection.
The only place that Dawkins' analogy fails is that it requires a bit more abstract reasoning on the part of the reader and those ideologically opposed to his point are often incapable of supplying it.
Brady Chase wrote:
"I am not blind, ignorant, angry, or unreasonable. "
You are certainly unreasonable, and you are blind in that you start with your conclusion and work backwards to your facts. If you believe in creationism, then you are ignorant, because such beliefs distort your perception of the entire universe.
I don't maintain, however, that you are stupid. Remove your Christianity and I have little doubt that you would find evolutionary theory persuasive. Ideology is a bitch.
Dawkins' analogy is actually even more favourable to natural selection. His “Methinks it is like a weasel" specifies one and only one outcome (plus eventually a few variations). Natural selection doesn't care about the outcome (the target phrase) - many outcomes are possible. So if his analogy with one specific outcome works natural selection with many unspecified outcomes certainly must work.
Brady Chase's argument actually reinforces the case for Evolution Theory.
Sorry, I'm out of shape today. It's David Evans' argument, not Brady Chase's.
Seems like the analog for the predetermined phrase is the dictates of the environment an organism lives in. It's not as concrete, simple, codified as the english phrase "methinks it is a weasel" but every environment clearly corresponds to optimal ways to survive in it -- that is the concrete predetermined "goal" selection is working toward. The environment is the template selection and mutation are working off of and it doesn't take any conscious calculation to refine organisms toward that end.
“He introduces his famous, “Methinks it is like a weasel” experiment. More precisely, he does two experiments. In the first, random strings of characters are generated until the weasel phrase is obtained… Of course, the first experiment fails completely to reach the target phrase...”
Probably because the odds of that randomly happening are more than about one in 26 to the 28th power, or 1 in 4,161,536,836,220,040,000,000,000,000,000,000,000,000.
“In the second he starts with a few randomly generated strings. Then the few strings that happen to have any resemblance to the target phrase, even if it is just having one letter in the right place, are used as the progenitors for the next generation of phrases, in a process modeled after genes mutating in the offspring of the current generation. This process continues until the target phrase is reached.”
What is the mechanism that chooses which strings to keep?
Whatever the mechanism is, doesn’t its choice require a predetermined goal, an intelligently determined design?
“Plainly, then, it is not knowing the target phrase that guarantees success. Rather, it was the presence of a consistent selection mechanism, coupled with copious variation, that led to success in the second experiment.”
Plainly wrong. See above.
“The question, then, is whether evolution by natural selection mimics the relevant parts of Dawkins’s experiment. Clearly, it does.”
The strawman constructed in the OP tears apart an argument against natural selection (which is non-random), whereas the criticism it claims to address is actually against mutation (which is random, as in non-directed). Busted.
Mutations occur. Yes
Mutations can take many forms - duplications, translocations, inversions, deletions, etc. Yes
Organisms can alter their mutation rate. Yes
Mutations are random with respect to fitness. Yes
Mutations can be beneficial - increase fitness. Yes
Need I go on?
Whatever the mechanism is, doesn’t its choice require a predetermined goal, an intelligently determined design?
Sort of, and Nope. And at least two posters addressed this before you wrote your post - did you read the comments at all? The environment provides the 'goals.' There can be many, they can shift, and no obviously things like oxygen pressure, temperature, or availability of one prey species over another doesn't require an intelligently determined design.
Think of 'target' as analogous to 'good fit to the local ecology.' Absolutely, the original weasel program is simplistic in terms of having a single unchanging target; in real life ecologies change and there may be more than one 'good fit.' But no, the concept that the target represents does not require any intelligent design.
To eric #17:
“The environment provides the ‘goals’… Think of ‘target’ as analogous to ‘good fit to the local ecology.’ Absolutely, the original weasel program is simplistic in terms of having a single unchanging target…”
The environment is kind of like an orchestra conductor.
He waves his wand and a reproductive system sounds.
(You need one of those, or the symphony’s over before it starts!)
Another wave of his wand and systems to identify, acquire, ingest, and digest appropriate nourishment blares.
(Again, these are pretty important to the ecology, else the ecology dies out pretty quickly.)
Another wand wave and locomotion’s lumbering notes are heard.
(Best to have wings, fins, legs - get some “exercise”, get out there and see that there ecology.)
A couple more subtle conductor wand waves…
and kidneys, bladders, brains, and eyes… and
I've suddenly remembered why I banned you the first time around. Bye!
... just listening quietly from the back ...
Was there not a recent article on ScienceBlogs which told of study which showed that people, of any stripe, would believe more firmly in their conviction after being shown actual, factual evidence to the contrary?
When I was a high school freshman a guy came to the school-1971, and laid out the random-parts- of- a-watch scenario. Wondered why the school would let the idiot in the door.
As an abstract process( algorithm ), selection by natural selection is quite simple:.
Step 1) Replicate generation n with noise
Step 2) Accept only those from generation n+1 that meet an acceptance criteria.
Step 3) If any n+1 survived, go to step 1
In the end, I always go back to the simple tripartite division of evolution:
1) The fact of evolution,
2) The path(s) of evolution
3) The process(s) of evolution.
(1) is incontrovertible to every educated person on plant earth today...at least to those who do not have a cynical contempt for humanity's ability to understand the world.
(2) is a random process requiring empirical investigation to establish. Getting it right means solving a complicated puzzle with limited data. Every new piece of data requires revising the story as more pieces of the puzzle are discovered.
(3) requires theoretical modeling and experimentation. Natural selection, sexual selection, etc are theoretical concepts that have large explanatory power much more than the creationist concepts which have been thoroughly falsified and discredited.
If the current state of evolutionary theory ever proves to be wrong or somehow woefully inadequate, going back to creationism or intelligent design is not ab option. That has been forever foreclosed to the educated portion of humanity that does not have a cynical contempt for humanity's capacity for discovering truth. Some other scientific theory would have to take its place.