Sunday, December 6, 2009

Darwin's theory is coherent and hypothesis generating? Can a theory be proven true?

Stanley Sessions responded to my comments at Nature Precedings:http://precedings.nature.com/documents/2887/version/1#comments
"I think I understand at least some of what you are getting at…wonderful! I think of the schizoid pickle physics is in right now compared to biology, the latter thanks to the remarkable coherency of evolutionary theory. However, you get into some dangerous territory when you speak of scientific theories “proven to be true”. You can do that in math, but isn’t that why mathematics is usually stuck in Humanities with philosophy rather than in the Sciences? Finally, to me the Theory of Evolution (like any good theory) is a good theory not so much because of what it explains, but to the degree that it is useful in generating testable hypotheses."

my reply:
I am not surprised that you would consider the existing evolutionary theory remarkably coherent. It is common for lay people and scientists not in the field or all casual believers to take that view, and I myself was among them just a few years back. But any disinterested professional who is familiar with the details would think otherwise. The devil is in the details. For a true theory, the more details you know, the more you would marvel at its beauty. The opposite is true for a false theory. So, it does not bode well for any theory to claim a lot of lay admirers. “When told that only 3 men in the world understood Relativity, Sir Eddington asked "I wonder who is the third?" Lay people and non-specialist scientists have no business in judging a science theory one way or the other. Evolution theory should not be simpler than relativity or easier to grasp for lay people. After all, it is evolution that created the brain for understanding relativity. A science that is so simple that a lay person could feel like he knows all about it, as is commonly seen in forums on evolution, can hardly be called science. The 3000 year old book I-ching is the most mysterious book ever written and only a handful of people the like of Congzi can really claim some understanding of it, and yet that has not prevented it from being the very foundation for the world view of the largest population on Earth, most of whom have no first hand understanding of it.

A coherent theory should not lead 100% lay people and 99% biologists to have the wrong answer to a most basic question that has been extensively studied for nearly half a century. Which, frog or human, is closer to fish in sequence similarity in any given gene? And yet the other side of the same question can be answered correctly by nearly all people, thanks to 150 years of teaching Darwinism. Which, frog or fish, is closer to human in sequence similarity in any given gene? If you don’t believe me, try these two questions on your biology colleagues. Being a casual believer of Darwinism until a few years ago, I, a professional biologist with 25 years of research experience, answered incorrectly a few years ago and found the right answer the hard way, by reinventing the wheel (no specialist has offered it in any books for the non-specialists and what is being offered can only mislead). That was truly a shocking experience and a big reason for me to get into studying evolution or to try to solve the puzzle myself given that few cared to do it. I have done enough to realize there is a big hole in our present state of affairs. The seeming coherency is a huge bubble. The fact that one must use Darwinism for phenotypes and the negation of it (neutral theory) for genotypes, despite the inseparable unity of genotypes and phenotypes, would easily strike any biology student as odd and incoherent.

The only reason that specialists seem to have a consensus view in public is because they have all agreed that contradictions are no big deal and that they can never be certain about anything in evolution, more for serve-serving than for the love of science/truth. They of course know it is a big deal in their private moment, as is apparent from the fact that you rarely see them teach the contradicting details to the non-specialists in books such as Why Evolution Is True. (While another person may easily read this as a conspiracy, I would just view it as simple human nature) One of the results of their selective teaching is that nearly all people know the answer to one side of a question while nearly no one knows the other side of the same question as mentioned above. In contrast, physicists, more of a hard core scientist, will not easily settle for any theory with contradictions, and hence no consensus yet on the cutting edge physics as you have correctly observed. But that is healthy for science and is what science is all about. Evolution research could for its own great benefit really use some challenge to the paradigm, especially a voice asking for a contradiction free theory and for not stretching a contradiction free theory of microevolution to where it meets contradictions and hence does not belong. No truth is afraid of a competitor because there can only be room for one truth, and there is no possibility of two theories explaining nature equally well.

My point is mostly about proving a theory wrong. It is debatable whether a theory cannot ever be proven true (see below). But it is often overlooked by biologists that a single contradiction/exception should prove a theory wrong. Your point about generating testable hypothesis is absolutely essential. But that is really a minimum, and meeting that standard is not enough. The theory must also be prepared to accept all the consequences of the test, one of which is to kill the theory or at least narrow its domain of relevance. If the theory stands regardless of the test results, then it is not a scientific theory. Unfortunately, the modern evolution theory is like that. I have personally found and published exceptions/contradictions to the theory and few cared. And many others have of course experienced the same. Few textbooks ever highlight those contradictions. And yet a sharp reader can still infer from what is not said. What he does not read anywhere is a statement by a reputed evolutionist that the modern evolution theory has no contradictions. But I do not waste energy to complain and I am fully aware just like any good scientists that a theory is never overthrown by contradictions but must be pushed away by an alternative better theory. Scientists who overlook those contradictions do it at their own peril as they are wasting their careers by barking up the wrong tree. There will always be people though in small number who will pay attention to those contradictions and try to do something about them and hence have a chance to make history.

For theory testing to be meaningful, the standard of no single contradiction is essential. If you allow contradictions, why bother testing in the first place? No doubt the modern evolution theory has been tested true or generated testable hypotheses but of course only in microevolution (e.g., Lenski’s and Grants’ experiments on bacteria and finches). It is equally clear that it is largely irrelevant to macroevolution if we use the same standard as we use for claiming it true for microevolution. All the contradictions are in macroevolution and none in microevolution.

A theory that has not even explained all known relevant facts is not even worth testing because it is already false. It has already allowed exceptions, which makes testing meaningless. To explain all existing relevant facts is therefore a minimum for any theory to qualify to be tested. And if a theory has explained all relevant facts, it would be impossible for it not to be able to generate testable hypotheses that would just confirm it further. Explaining a fact is the same as predicting it when a theory is axiom based rather than induced from factual observations. If a theory is induced from 10 known facts, then none of these 10 can serve as evidence for it (to do so is a tautology) and further testing is necessary. In contrast, if a theory is deduced from an axiom, then it is self evidently true and can exist independent of any facts and can rightfully claim all facts as evidence if it deduces/predicts/explains all of them.

I here provide a simple rationale for why a theory that has explained all known key facts without a single contradiction must be true or has been proven true. We first grant that for any domain of nature there is only one unique true theory or law that explains or governs it. This is at least 100% true for all the known laws of nature. If a theory has explained all key facts without contradiction, it is effectively indistinguishable from an ideal true law. Since there cannot be two true laws or more than one true law for any single domain, any theory that explains all key facts is effectively the true law. What are key facts? A key fact is any representative of a key class of facts that have the same fundamental feature. The knowledge of today cannot possibly have all the facts of a class known, which may be infinite. But it is not unrealistic for today’s knowledge to include at least one fact out of every major key class of facts of certain domains. And explaining one such representative of a key class is of course equivalent to explaining that whole class. In most cases, to explain all the known key classes in a coherent fashion (while granting the existence of some unknown key classes) is simply impossible unless one has the true law, given that all facts are connected by a single law and cannot possibly be connected by any laws other than the single unique true law.

Have Newton’s laws of motion been tested by all possible tests? Of course not, impossible to do. Are they proven laws? Of course they are because they explained every key fact known even if we grant there are unknown key facts and tested true in every test within their domain of relevance. It is also important that they are axiom based or self-evidently true, which means that their truthfulness is only a matter of how general or broad of their domain of relevance rather than having a possibility of being not true at all in any domain. Are they also wrong? Yes, outside of their domain of relevance.

The nearly half century old genetic equidistance result of Margoliash for cytochrome c is a key fact that is representative of most genes of most species. It is the answer to the above simple question on frog/fish that almost all people fail. How this single fact is interpreted determines how modern evolution research is to proceed and how half of the record of evolution (the DNA half vs the other half fossils/phenotypes) is to be interpreted. Its incorrect and tautological interpretation, the molecular clock and the neutral theory, has misled the field of evolution for nearly half a century with grave consequences. For one, no non-specialists know much about it and the less they know it, the more coherent they view Darwinism, which suits the Darwin specialists just fine. Some actual data for this. The wiki page on Motoo Kimura has a pathetic ~1500 views compared to Darwin with ~260, 000 views a month, never mind that it is Kimura’s theory rather than Darwin’s that is most relevant to modern evolution research on DNAs in the past half century. Also, the page on the neutral theory has 3000 while natural selection has 90,000 views a month.

The other grave consequence is to misidentify chimpanzee as our closest blood relative among non-humans in total violation of the fossil record and common sense. Another is to give us endless conflicting interpretations on the phylogenetic position of some species such as tarsiers. A set of conflicting interpretations must contain a false one. Any theory that can turn perfectly solid DNA data into a false interpretation of reality can only be incorrect, regardless whether it sometimes can also produce a correct interpretation by accident. Any theory that has turned DNA data into conflicting interpretations of the reality of phylogeny has of course self-proven itself false.

The fact that Darwin’s theory can coexist in peace with such a false theory of molecular evolution in a form termed the modern evolution theory does not bode well for it. All these theories are from the beginning based on population genetics and should have just stayed where they really belong, microevolution and population genetics. As far as I am concerned, no theory of any kind, except the maximum genetic diversity hypothesis, has correctly explained the equidistance result and thus has any chance of being true for the whole domain of evolution or has even qualified as worth further testing.

No comments: