Perform Your Own Darwinian Evolution Experiments At Home!

Darwinists like to remind us that it’s chance AND the necessity of selection, but it seems to me that they don’t really appreciate how that every step of the way must be “sheer dumb luck” before it can be selected.

Sure, when that first mutation comes along it's pure dumb luck. But selection (allegedly) preserves it if it in any way contributes to a difference in reproductive success, and that way when the second pure happenstance mutation comes along you get to add it to the one already existing, and this brings about a more probable state of affairs than if they had both just happened by pure chance at the same time, and therefore the improbable becomes probable, and the probable certain, and therefore evolution is a fact. Mung

Mung brings to mind another problem with RM + NS. The good mutation must occur in the embryonic cell and it must be a dominant trait if it is to be passed on to another generation--unless two descendants with the recessive traits find each other and mate. It's like drawing the Q in Scrabble and never getting to use the U. :( kathy

Keep in mind that natural selection doesn’t work like a spelling checker. It doesn’t turn a bad mutation into a good one; it just throws away bad mutations.

Actually, this is a bit misleading. Natural selection no more throws away deleterious mutations in a single generation than it fixes a beneficial mutation in a single generation. As such, "bad" mutations tend to accumulate, certainly at a greater rate than "good" ones (they happen more often). This creates an additional problem that must be overcome even if a "good" mutation does happen to show up.

Mung

Darwinists like to remind us that it’s chance AND the necessity of selection, but it seems to me that they don’t really appreciate how that every step of the way must be “sheer dumb luck” before it can be selected. This means that every last minute Darwinian gradualist increment from sludge to Judge John E. Jones III had to be unmotivated. Selection only preserved what “sheer dumb luck” produced. Therefore it is interesting to see just how long it takes to stumble from “The quick brown fox jumped over the lazy dog” to “Jackdaws love my big sphinx of quartz”—this even without the requirement that each step be somehow advantageous. Darwinists make their computer simulations too easy on the randomizing side and too teological on the selection side. Rude

The text mutation generator is not really a simulation. It’s just a fun demonstration that offers an intuitive grasp of the impracticality of using random changes to do anything useful or meaningful in an information-encoding system. By the way, as DaveScot points out, codons include three base pairs. Since ACTG is a base-four system, there are 4^3 or 64 possible codon arrangements. That’s analogous to the 65 characters used in the written English language (upper and lower case alpha characters, plus numbers and punctuation). GilDodgen

JohnnyB: "“How does RM+NS work for micro evolution?” It doesn’t, generally." This depends on your definition of micro evolution. If micro evolution is defined, as it is traditionally done in creationist and some ID circles, as change within a species. (Species defined classically as groups that are genetically capable of breeding producing fertile offspring.) Then natural selection alone, without random mutation, is fully adequate to account for such variety. After all, animal breeders are able to produce astounding variety from basic wolf stock by breeding alone. Certainly natural selection is capable of producing this level of variety. I have yet to be convinced that random mutation ever does anything of significant value. bFast

Note -- a better simulator would also keep near-neutral mutations, such as substituting a K for a C and such things. Note that this will not necessarily help the simulator produce results and may actually hinder it. johnnyb

"How does RM+NS work for micro evolution?" It doesn't, generally. I'm sure there's some tiny minor circumstances where it might. However, RM is on the way out as the basis of selection. But somehow this proves Darwinism right instead of wrong. "Rejecting entirely random genetic variation as the substrate of genome evolution is not a refutation, but rather provides a deeper understanding, of the theory of natural selection of Darwin and Wallace." johnnyb

The major problem with this "experiment" is that the variables are too large. DNA is composed of 4 bases, A, C, G & T. Using an english sentence composed of 26 letters, 10 numbers, returns, puncutation marks and so on gives the program too much room for error. Certain bases only go with other bases, using all those others? That's cheating.

Actually coding genes use ACGT triplets called codons where each codon specifies one of 20 amino acids. 20 amino acids, 27 letters in the alphabet. Not a bad match at all. Try again. -ds Jeffery Keown

How does RM+NS work for micro evolution? Barrett1

Cousin Gil, You said: "Keep in mind that natural selection doesn’t work like a spelling checker. It doesn’t turn a bad mutation into a good one; it just throws away bad mutations. You can simulate natural selection by clicking on the back button in your browser when a mutation messes up a word. No cheating is allowed. You can’t preserve a letter mutation that produces a non-word because in the future it can be used to convert one legitimate word into another. Natural selection has no foresight." Bingo. I've never heard this fact expressed better, or with a better analogy. Stu Harris www.theidbookstore.com StuartHarris

I think that the "spell checker" discussed in the article is one of the simple sort that simply says "this word is in the dictionary, this one is not" rather than the advanced sort that gives a list of possible misspellings, or does an auto-replace when the list has a single item in it. bFast

This simulation strikes me as silly for a few reasons. First, there's no selection half of the equation, it's just changing random letters and letting you look at how strange everything gets. Second, starting from a perfect sentence and then noting the only direction is down doesn't seem very productive or relevant. Third, there's no potential for neutral mutation (as a consequence of the second item). Fourth, you all are willing to claim Avida and the like prove nothing and this little applet proves everything.

Also, my sped-up version of this game managed to turn dog into man in only 10,000 tries. Clearly that proves evolution right there.

Actually it disproves evolution right there because man did not descend from dog. Present company excepted of course. -ds Tiax

It's my understanding that some DNA polymerases perform a proofreading activity to excise incorrect base pairs and insert correct ones. It's not the same as a spell checker, but error detection and correction schemes are a characteristic of well-designed information coding systems. dgw

On the subject of spelling checkers: Keep in mind that natural selection doesn't work like a spelling checker. It doesn't turn a bad mutation into a good one; it just throws away bad mutations. You can simulate natural selection by clicking on the back button in your browser when a mutation messes up a word. No cheating is allowed. You can't preserve a letter mutation that produces a non-word because in the future it can be used to convert one legitimate word into another. Natural selection has no foresight. The bottom line is this: The scientific controversy concerning the substantive creative powers of random mutation and natural selection in biology is over. If there is a fact concerning RM+NS as an explanation for all of life's diversity, complexity and information content, it is that this proposition is not only false, but spectacularly false. This "theory" -- which is sold to the benighted, unwashed masses as a foregone conclusion and a fact with the strength of the theory of general relativity -- doesn't even qualify as intelligent or reasonable speculation. It is incoherent, preposterous, hopelessly in opposition to simple statistical analysis, and just plain silly on the face of it. However, as I pointed out in a previous UD blog post, it will be defended to the death by Darwinists by any means available. To do otherwise would mean admitting that they are followers of a false religion. GilDodgen

"The evolution of language was indeed witnessed as it was happening, but so was biological evolution. Too bad the overhwelming majority of witnesses are dead. In the case of language evolution, beyond at most a few thousand years ago, there are no remains of early languages, no “fossile” languages." - Raevmo Maybe I've misunderstood you, but there are indeed living witnesses to language evolution. Language evolves continuously. In my own lifetime, the meaning of expressions like "begging the question" (it's increasingly used to mean "raises the question"), "gay", "queer", etc. have changed. But as others have noted, language doesn't change randomly or mindlessly. Humans assign new meanings to words or create new words and grammatical constructions because they find it useful and beneficial to do so. russ

But we learn from this example that the single-letter mutation approach is inefficinet in achieving the end goal of black fox. In fact, the experiment shows that effective tuning takes place at a context level (is the fox brown, black, red, tan, or white?) rather than at a symbol level. Tweaking the symbols in a sentence is analogous to changing the object code in a computer program to improve its performance. (How many single event upsets have a positive effect?) What this experiment also implies is that the purpose of the design becomes important--choosing a fox color that generates the most clicks. For typing or font selection, one would revert to the original sentence: The quick brown fox jumps over the lazy dog. A designer looking for alternative ways of displaying all letters in the alphabet might choose the less familiar "Jackdaws love my big sphinx of quartz." Choice of fox color has no value in this case. If we assume something is designed, the question that follows is: how is the design adjusted? dgw

Reamvo, you are suggesting that the evolution of language and the evolution of organisms have a tight paralell. You are also suggesting that the example provided is somehow nullified by this fact. So I believe that you are suggesting that the conversion of, "The quick brown fox jumped over the lazy dog" to "The ferocious black fox leaped over the sleeping dog" is an example of the evolution of languages. Of course, it isn't. This is an example of the evolution of information within the context of a single language. While there may be some differences in the actual language of DNA between the different domains of life, the primary difference between one animal and the other is found not in a change of the language of DNA, but a change in the information within the same language. Therefore, your argument does not nullify the example provided. bFast

Populations evolve, not individuals.

Assume a pond filled with a population of only 10^14 such phrases (the number of prokaryotes living in the typical human gut). Assume a mutation once every 10^7 replications, or about 10^7 mutations per generation.

The total of number of possible recombinations and point-mutations is on the order of the length of the sequence raised to the fourth, or ~10^7. That means that almost every possible self-insertion and point-mutation will (likely) be tried in the first generation or two.

Now let's consider our specific sequence, "The quick brown fox jumped over the lazy dog". One possible point mutation, as already pointed out would be "The quick brawn fox jumped over the lazy dog". And this will be tried each generation. But so will these, "The dog jumped over the lazy dog", or "The quick brown lazy dog", "The quick brown fox jumped over the lazy brown dog." or simply "The quick brow", or "fox jump".

Assuming equal selection (drift), the pond becomes full of variation. And with a suitable selection mechanism, there is little limit to what might occur. Testing reveals that even a few thousand generations will yield a vast variety of phrases, many of considerable length (depending on the selection criteria).

Mutations occur in individuals don't they? If mutations are the source of change in evolution then it's the individual not the population where evolution happens. Are you saying that mutations don't occur exclusively and instantaneously ONLY in individual organisms? -ds Zachriel

I didn't *really* wonder why there is less controversy in language evolution. An obviously failed attempt at mild sarcasm.
The game of randomly mutating words shows the absurdity of thinking that RM+NS might create something meaningful. This suggests to me that the word game is a poor analogy of RM+NS, but it also made me wonder what really drives language evolution. Clearly there are some elements of intelligent design (like inventing new words or abbreviating older words), but it seems to me there are also lots of random mutations. How did such mutations spread?

The evolution of language was indeed witnessed as it was happening, but so was biological evolution. Too bad the overhwelming majority of witnesses are dead. In the case of language evolution, beyond at most a few thousand years ago, there are no remains of early languages, no "fossile" languages. Yet purely based on similarities between current languages it is concluded that there must have been a common ancestral language (say to the indo-european languages), without any direct evidence.

Evolution has never been witnessed. Before humans were around there was nothing that could serve as a witness and during recorded history there has not been a new species seen to emerge. You're going back to moderation and I suggest you start putting a little more thought into what you write if you expect your comments to appear. -ds Raevmo

Raevmo -- This where the debate becomes surreal. Language trees can be developed and traced, that is true, but why would you use the development of language -- indisputably linked to intelligence and design -- to make a case for the evolution of biological unless you are making the case that life developed via intelligence and design? tribune7

Actually, languages do evolve by mutation, selection and drift. In fact, the evolution of languages can be mapped on "phylogenetic trees" in very much the same way, using very similar statistical tools, as the evolution of species can be inferred from analysing differences in their DNA "language". Unlike biological evolution though, there seems to be far less controversy over language evolution. I wonder why.

I'm beginning to wonder why I took you out of moderation. 1) The evolution of language was witnessed as it was happening. 2) Language didn't evolve from lifeless chemicals into a hideously complex world of self-replicating molecular machines. Now you don't have to wonder why any more. -ds Raevmo

Continuing, if single-letter mutations occur for the daughter sentence the quick brawn fox, the spell-check correction repairs the mutation 42% of the time, and reverts to the orignal brown-fox sentence 12% of the time. The new variations that occur are braw at a frequency of 12% and bran at a frequency of 10%. Other words like brawl occur less frequently. Accordingly, the probability of proceeding from brown to brawn to bran is about 0.5%. If one assumes a 5% probability for each generation, then the transformation from brown to black has one chance in 3.2 million of occuring via this 5-generation pathway. This evolutionary procedure is more optimistic than the author's approach. However, a free-market system is still more likely to favor daughter sentences that use colors. It is interesting to note that spell-check-corrected single-letter mutations produce wordlists within a short spelling distance, but on a contextual level, the separation is far greater. Mismatches of this kind could be a clue that enables discrimination between outcomes that were generated from the bottom up and outcomes that were designed at a higher level. dgw

If brown has a single letter error, including a deleted letter or an appended letter, there are 162 possible mutations. Using a spell-checker to repair the mutation provides a powerful correction mechanism. 52% of the time, the letter error is corrected and brown is retained in the sentence. 14% of the time, the spell-checker chooses brow instead of brown. 8% of the time, brawn is selected. Other words appear less frequently. dgw

Cousin Gil, About halfway through William's page on random generation he has: "There's an Even Bigger Problem Let's say you want the word Brown to evolve into the word Black. Shouldn't be too hard, should it? Only four letters need to change after all. But even if you could get the mutations to concentrate just on those four letters, you'd still get a mis-spelled word, which natural selection would eliminate before it ever evolved into the correct spelling." There is quite obviously an even BIGGER problem than what he states here. In the Darwinian process, there is no concept of "wanting" with foresight to evolve one thing into some other distant pre-defined thing. If you are setting up a process "to get the mutations to concentrate" on any desired outcome, then you have a teleological process, not a Darwinian one. ALL the supposed simulations (Avida, Nilsson/Pelger, etc.) that purport to demonstrate how a Darwinian process can create CSI have this problem. Stu Harris www.theidbookstore.com StuartHarris

Interesting concept. One modification might be to consider pathways that allow a chain of mutations that survive a simple spell checker. If the rules support the addition, deletion, or change of a single letter, then brown can mutate to black via a pathway such as: brown to brawn to bran to brand to bland to blank to black. There are perhaps many such pathways, some possibly shorter. It would be more difficult if the rules required that mutations meet both a spell-check and a grammar check. Also, if the mutating sentence can self-check, it's possible that some mutations would be close enough to correctly spelled words to survive after correction. dgw