Perhaps nothing, and this may indeed be a "ghost in the gaps." If it can be eventually shown exactly how human consciousness is produced from neural interactions, to the point where any "ghost" is completely unnecessary, than to posit a soul is pretty pointless. We aren't there yet, and it is still possible that the "ghost" is responsible for driving the machine, but there is no evidence for this.
I do not believe that one needs to understand exactly how brains generate consciousness in order to establish that that is what they do
It isn't so much a question that the neural machinery is responsible for "generating" consciousness. We have known since Phineas Gage and others that brain damage can fundamentally alter who a person is, and we now know a great deal about where memory and learning take place, how long-term potentiation makes simple learning possible, how messages are passed back and forth via action potentials, etc. The question still is one of how all this electrical and chemical activity in summation amounts to a human being. And until we know that, while there is no evidence for it, the answer could still be the "ghost" organizing neural activities in particular ways.
I used to be a bigger fan of Kuhn's, but I now believe that paradigm shifts are less revolutionary and more incremental than he portrayed.
I don't buy Kuhn's argument completely by any means. I do believe that science actually progresses, and does not simply go from one paradigm to another. What is interesting to me is the tendency for a paradigm to dominate scientific research in an area and try to explain outliers either within that paradigm or as exceptions, until finally enough research is conducted as to make the old paradigm meritless. A good example of this for me is the biomedical model of mental illness dominating psychiatry.
Francis Collins, a devout Catholic,
I should point out that Francis Collins was an atheist and converted after his work in science. His examination of the genome led him to a belief in a creator. The design appeared too "designed."
We haven't proven abiogenesis, so the "explanation" has shifted to the gaps that science has yet to fill.
Collins and others are loathe to look for "god in the gaps" which are (as you know) failures to explain particular events. Collins, for example, points out that while there are certain things we can't explain in the evolutionary record, this doesn't mean we won't be able to. Guided evolution is often a "god in the gaps" argument, and when it isn't it is superflous. However, the strong versions of the anthropic principle are more concerned with the fine-tuning of the universe itself.
As Dawkins has pointed out, the "fine tuning" argument actually works against creationism, as it can be explained in terms of natural evolution.
The fine-tuning arguments go WELL beyond evolution. They concern far more things the laws of physics in the universe itself, and the fine-tuning necessary just to have an earth CAPABLE of supporting life. Compared to those kinds of arguments of fine-tuning, the "tuning" necessary for evolutionary changes is nothing.
Besides, there are perfectly good non-religious explanations of why human beings would come to believe in gods.
Yes.
I know of no such cases, but I am willing to be convinced.
There is a shrine of Lourdes in France where numerous miracles are claimed to have occured. So many, in fact, that a there is a medical bureau made up at times to examine these claims. The bureau is composed of doctors of with and without religious affiliations. When they determine that a so-called miracle can't be explained, they pass it on to the International Medical Committee in Paris, for further examination. If they can't explain it, they call it "medically inexplicable." It could be as simple as we just don't know what happened, but the point is these are documented cases examined by multiple experts under great scrutiny and the answer isn't there.