Any experiment needs to be risky. That is, the experiment should be able to
equally test the hypothesis and the null hypothesis. From my own experience, experimental controls can sometimes take up 80% of the samples in any given experiment. Those controls are there to see if the null hypothesis is correct.
Nowhere do I see ID experiments that test the null hypothesis. For example, JBR suggests that beneficial mutations that occur on bacterial plasmids are due to an intelligent designer. So what is the null hypothesis? I would think that the null hypothesis would be beneficial mutations occuring within the non-plasmid genome, and the literature is full of such examples. So does this mean that non-plasmid beneficial mutations falsify ID? I doubt it. JBR and others will probably claim that these are designed as well. That is certainly what Behe claimed when Hall observed the production of a novel beta-galactosidase gene that occurred within the non-plasmid genome in E. coli (
reference).
The other argument is that experiments which fail to produce evolved structures are evidence for ID. Therefore, such experiments can count as ID experiments. Nothing could be further from the truth. Any theory must stand on it's own, and it has been shown that ID can not. As an example, the precession in Mercury's orbit that was not predicted by Newton's Laws which told us that there was something wrong with those laws. However, the falsification of Newton's Laws under certain conditions did not prove the existence of supernatural forces moving Mercury about the Sun. Even if the theory of evolution is proven false it does not evidence the accuracy of ID in any way.
Edited by Taq, : No reason given.