Along crashfrogs reasoning, say every spider of a certain species in a genetic "island" builds its web in the same way, couldn't that be evidence that the spider is following a "program" that doesn't contain a conscious component?
This depends on what "in the same way" and "conscious component" mean. Those terms are not easily defined.
If the spider is following a "program", then I doubt that it is the same type of program that we use with computers. That is, I doubt that it is a fixed sequence of mechanical steps. It seems to me that it would need a program that involves feedback and adjusting what it is doing based on the feedback - a sort of self-adaptive behavior.
If the spider hatches, has no exposure to other spiders building webs, and then builds a web exactly conforming to other webs in a given sample group of spidrs of the same species, wouldn't this be strong evidence of genetic rather than consciously driven behavior?
When I suggested in an earlier reply, that a bird has to learn how to build a mouse, I did not intend that to imply learning by imitation (or copying). I meant only learning by trial and error discovery of what works. In the case of the spider, I would guess there is little learning. That is, building the second web probably does not build on what could have been learned from building the first web. For the bird, I would guess that there is some learning, so that building the second nest is probably influenced by the experience of building the first nest.
Note that I am not a biologist. I suspect some of the above could be tested empirically, and perhaps this is already known.
Fundamentalism - the anti-American, anti-Christian branch of American Christianity