Monday, August 25, 2008

Well, now the nameless one says his question was about cytochrome-c test for evolution.

     Interestingly, I can't find that question. Of course my browser is acting up at the moment and cutting blog posts with a lot of comments. So, I decided to check google and see if it had anything to say about a cytochrome-c test for evolution. First off, I am suspicious of the 10^93 number. Oh, I am sure that Hubert Yockey has computed what he believes to be a real probability. But then, I believe the creationists are sincere when they trot out their "probabilities" too.
     The big problem that I have with the claimed tests in general is that they are of a "confirm or inconclusive" form. If the data they take as confirmation were not there, it would not be disconfirmation. The same appears (to me) to apply to this test. If cytochrome-c were not found or were vastly different between chimpanzees and us, I don't see how that would constitute negative data. I believe large-scale evolution is plausible (though I don't think testable) and I don't see how that would have changed if that outcome had been different.
     Something is a little more problematic for this particular test. Talkorigins reports that the cytochrome-c sequences in chimpanzees and humans are identical. Now, for the given probability to be meaningful, there can't be any selective pressures on those sequences. If there are selective pressures, then small-scale evolution (which is really a direct observation) will tend to lock them in and so the chances would not be so long. But if there aren't any such selective pressures, the sequences should have become different through mutation. We are talking millions of years and an imperfect replication process.
     This test looks like an anomoly even if large-scale evolution is true. I don't see it being taken seriously unless people are blind to problems in the testing. Under the large-scale evolution hypothesis, human-chimpanzee similarities in cytochrome-c should indeed be greater than human-yeast similarities. But an identical form should mean either that there was not a branching off point millions of years ago, or that a selective pressure favors that particular form (in which case, the stated probability is not valid.)

4 comments:

zilch said...

The big problem that I have with the claimed tests in general is that they are of a "confirm or inconclusive" form. If the data they take as confirmation were not there, it would not be disconfirmation. The same appears (to me) to apply to this test. If cytochrome-c were not found or were vastly different between chimpanzees and us, I don't see how that would constitute negative data.

If there were no overall picture of congruency between the cytochrome-c tree, and the tree of life as given by other measures, then you would be right: cytochrome-c relatedness would simply be deemed inconclusive. However, it is not just chimps and humans that show relatedness: the whole tree of life can be built just using cytochrome-c relatedness, with only a few anomalies when compared to the tree as constructed by, say, comparative anatomy. The chances of this congruence, not between one pair of organisms, but among several dozen (to date- more are added all the time) happening by chance are vanishingly small. How do you explain this "coincidence", if not by evolution?

This test looks like an anomoly even if large-scale evolution is true. I don't see it being taken seriously unless people are blind to problems in the testing. Under the large-scale evolution hypothesis, human-chimpanzee similarities in cytochrome-c should indeed be greater than human-yeast similarities. But an identical form should mean either that there was not a branching off point millions of years ago, or that a selective pressure favors that particular form (in which case, the stated probability is not valid.)

Can you give me some data to support your assertion that the identity of human and chimp cytochrome-c is an "anomaly"? The speed of genetic drift can only be given a statistical value, and five million years or so is not a great deal of time. Again, you seem obsessed with details to the point of ignoring the big picture. And claiming that this "anomaly" is taken seriously because people are "blind to problems in the testing" is, well, arrogant, unless you can show that this is should indeed be considered an "anomaly" that renders the whole rest of the comparison meaningless. I'd love to hear your demonstration of this, based on your knowledge of rates of mutation, the genetics of cytochrome-c, and statistics.

Given the complexity and imperfect record of evolution, we can never expect to have a perfect picture of what happened, or a perfect fit of data with theory. You are thus correct in saying that inconclusive data is not considered a falsification. What you ignore, however, is the vast amount of data which does show a pattern, which can be explained by evolution, but not by any other hypothesis so far. You are not seeing the forest all around you, but rather complaining about one or two missing trees.

Pvblivs said...

Zilch:

     I am coming from a "don't know" position. I am not advocating "origins come from X." Therefore, I need not explain any phenomenon.
     I am having difficulty finding a direct source on the rate of genetic "copying errors." I am only finding that there is a high degree of variability. At any rate, I will concede that my expectation that there should be some difference if there is no selective pressure on the mutations is intuitive.
     "What you ignore, however, is the vast amount of data which does show a pattern, which can be explained by evolution, but not by any other hypothesis so far."
     Where are all these failed hypotheses. If scientists have attempted numerous alternative explanations, I am not aware of it. But if I can see that, I will gladly admit my error. The only thing I am aware of that even pretends to be an "other hypothesis" is the old "godidit" claim. Since "godidit" doesn't explain anything (it makes no predictions of any kind) the absence of a hypothesis that explains the patterns is not surprising. But if there were other hypotheses, built on the pre-existing data, that fell as new data were collected, I will be very interested to learn of them.
     My perceptionis that scientists directly observed small-scale evolution, postulated large-scale evolution as an extension of it, and looked for data to confirm their hypothesis. I do not see where anyone has attempted to propose an alternative (other than the blatently useless "godidit") or looked for data that evolution would say should not be there. It's like your bragging about winning a race, but the only competition I saw was a housecat that seemed more interested in napping.
     Now, I could be wrong. Perhaps there were alternatives proposed. If there were, I simply haven't heard of them. I can certainly draw an erroneous conclusion when I am missing important evidence. But I see similarities between the responses and defense of religious belief. An obvious one is calling on me to explain things I never claimed to be able to explain.

Igor said...

One alternative hypothesis which did not survive is Lamarckian Evolution. The other I can think of was Lysenkoism, which turned out to be a an ideologically motivated hypothesis, which only became popular because it adhered to socialist principles. It is worth noting, that such blind adherence to that farce of a scientific theory set Russian genetics programs back 20 years.

Pvblivs said...

Igor:

     That is certainly interesting; but it is an alertnative within large-scale evolution, not an alternative to large-scale evolution. I have already stipulated that genes have already been adequately tested. I accept as theory that genes are at one thing responsible for small-scale evolution (which has been directly observed) and if large-scale evolution is accurate, then genes are likely involved in that too.