hrothgar, on 2011-March-04, 13:46, said:
Predictive Power
The examples that I am most familiar with are related to searching for specific types of fossils. Biologists have made specific predictions that transitional forms of type [X,Y,Z] should be found in specific types of strata. They have even launched (successful) expeditions to search for a given type of fossil in a specific type of location based on prediction.
Key Elements have yet to be tested
How is in any way, shape or form specific to the Theory of Evolution?
Falsification
First and foremost, we have experimental evidence that shows the emergence of a new species
http://www.newscient...in-the-lab.html
Try to be civil hrothgar.
Regarding your points: I knew about those expeditions, I just don't consider them to be predictive in the proper sense. Seeing that species A and species B look quite alike, and inferring that there is a species in-between A and B is only support for a much weaker claim that the development of species is continuous. Even intelligent design advocates don't question that. To be predictive, I would say that it must predict something that has not happened yet, not merely infer a missing link in what has already happened. The jury is out on whether evolution will have this kind of power, in the sense of making sensible predictions on the variations in the genome of a given population. Generally it is expected that if you reset the evolutionary clock, you will get a widely different world from this one, although I not recently that this has been disputed. Myself, I believe that evolution will have some degree of convergence, but at any rate, it will be virtually impossible for evolution to make the type of strong predictions that Karl Popper would like to see, at least in the lifetime of individuals. Even in the case of the article you gave, all of the populations that did not develop cit- had significant deviations in fitness - i.e. populations held in identical situations will tend to diverge rather than converge. At least in the short term.
As to the key elements that have not been tested, I was referring mostly to speciation. The article you cited most definitely does not refer to the emergence of a new species, as you claimed, but only to the emergence of a single new gene. Generally, to change species (in bacteria), you must make changes to at least some of the genes that are crucial to a cells functioning. This is very difficult, as most changes to most of these crucial genes will result in cell death after even a single mutation. In a sense, this is why species are stable in the fossil record.
The article you gave is very interesting, I had read other of their papers (where I got the 20000 generations of E.coli from) but I did not realise that they had discovered the emergence of a new gene. Let us be clear that this is Big News. It is the first example of a new and functioning gene with a new ability. My guess is that they are playing it low key until they can sequence the gene, as they will be worried that it might only end up meaning that E.Coli had this ability in the past, and "re-discovered it" rather than created it from scratch. (Obviously, if they had a gene that was different by only a few mutations that "broke" it in the past, it is much more easily re-discovered). I could probably do a literature search and find out, but I'm tired.
Finally, the article actually highlights the problem of evolution in complex species. According to the article, the rate of mutation in the genome is 10^-10 per cell per generation. And that is for a single mutation. Their populations were typically around 10^12, so they got a large number of mutations per generation, and it still took 35000 generations for the first example of a functional new gene. (According to their notes, all other phenotypic differences were due to changes in gene expression, not new genes per se). In general only a small fraction of the cells in a complex organism are involved in reproduction, eg in a female, mutations only matter in one cell per month, under the tenets of current evolutionary theory, only mutations in the egg can be passed on, so only they matter. Indeed, it is my understanding (although I'm not to clear on this) that all female eggs are present at birth, and they do not undergo cell reproduction, but sit with their dna mostly inert until they start to develop a few each month for the menstrual cycle. At any rate, only a tiny fraction of the cells in a higher organism have the chance to pass on mutations to the next generation, and ultimately, since only two cells are involved in reproduction, the number of cells in which mutations can be inherited is twice the number of humans born. Given cell mutation rates quoted from your paper, that implies that for a single mutation in a single gene, there would have to be 10^10 humans born, for a complete functional gene, 10^13 is a reasonable estimate. I.e., if we extrapolate the results of that paper to a human genome then we would have to infer that in the entire span of human existence, we have not created a single functional new gene. (The total number of all humans who have ever lived is definitely below 10^11 (100bn).)
Of course, this is not a problem for natural selection per se, because as previously alluded to, nearly all natural selection makes genes that already exist more prominent. In fact, many recent finds have suggested that natural selection is way more efficient than was previously believed, mostly because it is starting to appear likely that humans can "program" their offspring, via the inclusion of RNA in their reproductive cells (think its called epigenetic selection?). The exact how is still a bit of a mystery, but statistically significant differences in children have been correlated with information on the parents, like Stress, or Obesity: hence it is starting to look increasingly likely that RNA might be involved in "programmable" selection.
Since that was a bit of a long post, I will summarise it as follows:
(1) Evolution will always lack predictive power in a strict sense, because it will generally not repeat itself even in identical conditions.
(2) Speciation has never been observed in progress in the natural world. It is entirely unclear how a species can evolve a different number of chromosomes, since in it is believed that two individuals with different number of chromosomes cannot breed successfully, no matter how alike their genome. The chance of two individuals simultaneously evolving a mutation that complex is effectively zero.
(3) Natural Selection is probably more complicated/efficient than we currently believe. This would also make more sense of the vast amount of apparently unused DNA that we carry around - perhaps in stressful conditions we can start expressing some of these genes in our children to improve pheno-typic diversity.
(4) Current understanding of gene creation via random processes does not produce enough new genes to account for the differentiation of complex species. Either virtually all the diverse genes present in higher species were present in bacteria prior to the ascent up the evolutionary tree, or we are capable of far more rapid mutation by an as yet unspecified mechanism.
Finally, I will add that even among mainstream evolutionary biologists, there is deep dispute about the relative importance of various things. E.g. Multi-level selection vs selfish gene type models.
I think there remain too many unsolved problems in evolutionary biology to put it in the top tier of scientific achievement. Obviously, being a physicist, I am biased, but there is an economy and elegance in today's theoretical physics that is unmatched in any other scientific discipline. Using elementary field theory methods, I could, for example, prove that the sky is blue, better yet, I could get the exact scattering dependence of the atmosphere on the wavelength of light. For an encore, I could use some basic QM mechanics to compute the relation between refractive index and pressure, and compute the precise value of the proportionality constant. If I compared to measurement I would be better than 1% (on a dry day - humidity interferes with my measurement). All this requires only a double handful of postulates and one experimentally measured quantity (the fine structure constant).
Now if only elegance===usefulness
PS: People seem to be putting opinions in my mouth, so for clarification:
(1) I am not anti-evolution, I think it defiantly happened, I just think that current biologists faith in the details of the mechanism is misplaced.
(2) I am not against teaching evolution in schools. I just think its more important that parents should decide what their children learn. The danger of a tyranny of the majority leading to virtual indoctrination in schools is very real. It has happened before, and it doubtless will happen again.
The physics is theoretical, but the fun is real. - Sheldon Cooper