I believe strongly in the development of evolutionary science, and I have new hypotheses that ought to be tested that don't contradict any experiments done, nor do they contradict the overall modern evolutionary synthesis. However, these hypotheses, if proven, may open up some arguments thought long ago settled.
The role of random mutations is well researched. These are well observed and are thought to be the basic unit of truly new, potentially beneficial adaptations.
DNA repair Is a process where damage to the DNA is fixed. This process is imperfect, however and failure in repairing mutations is a variable source of mutations. The process of DNA repair is blind to the function of the gene it is repairing. Thus, resulting mutations tend be in random places, and the build up of mutations on sections of DNA that don't change the function of the gene happen at a well defined rate. Thus, a process of feedback of information whether the gene is functioning properly, is the only way to account for genetic drift NOT occuring on important functional elements of the gene. Natural selection acting on whole organisms holding that gene is the only feedback commonly agreed on by evolutionary biologists to occur. However, my view is, that with a hierarchy of genes (ie. genes that control a bank of genes, that each control a set of basic genes that relate to a phenotype), there needs to be a hierarchy of feedback to ensure that the lower level genes are functioning. Ie. there needs to be a form of selection within an organism such that each basic gene can be selected or rejected based on its function independently of all other genes. The corollary being that lower level genes that are suppressed in some way by some higher level genetic action, can have its function ensured with the same process of feedback.
There are a number of mechanisms that could be at play to ensure functional integrity of genes, that are not natural selection between whole organisms. One of the mechanisms proposed, which is quite likely to be involved in some way is through the selective properties of sperm. In this mechanism, sperm act as selective proxies for the organism but specifically for genes lower down in the genetic hierarchy. Thus, every lower level gene affects the selective aspects of sperm, and that which gets to reproduce has fully functioning lower level genetics.
In this sense, this hypothesis is not concerned with the genetic variability due to the shuffling of phenotypes that follow the laws of Mendelian Inheritance, but only the functionality of the individual allele itself. This hypothesis takes it as a given that the spread of mendelian traits is the primary source of variability in phenotypes that are subject to selection in a standard Darwinian way. This hypothesis is concerned with: 1)How "improved" versions of alleles arise.
2) How stress triggers greater mutations.
3) How latent phenotypes not visible in a species can become common again.
4) How many "truly new" genes are involved in speciation, and how many genes are latent ones that are re-activated (or de-activated), or inserted via horizontal gene transfer, or are just a previously unobserved combination of mendelian and non-mendelian traits.