Darwin Unhinged: The Bugs in Evolution

One big problem, I think, is with instinctual knowledge - let's say a fear of snakes. The problem is that conventional neuroscience tells us that groups of neurons develop concepts - such a snakes, chairs, cars, etc in a statistical fashion after processing many situations containing these items. The problem is that this would mean that no two person's symbols would be the same in internal structure. How then can any instinct that refers to such symbols be represented?

Humans are less driven by instinct, except perhaps as regards courtship and sex. Here sexual ideas seem woven into the rest of our lives in complex ways. You can't separate a part of cognition that might come pre-formed, from the rest of our cognition.

This is one reason I support far more radical alternatives - more like Rupert Sheldrake's morphic fields.

David

Yes. I think the problem is also that the neo-Darwinian evolution of the instinct would seem to require that just the right random mutations occur to the DNA encoding of the neural net. These random mutations would need to just happen to encode the particular fear-induced changes to the neural net structure. The physical changes to the neural net induced by the particular fear are probably very complicated, and the overall structure of the neural net even more complicated by many orders of magnitude. The brain is by far the most complex organ in the body. So it's the probabilistic problem again, the combinatorial explosion problem. The likelihood of this happening seems to approach zero because of the vastly greater probability of random changes being some other changes to the vastly complex detailed brain structure. It's not just any possible automatic instinctual drive, it's this particular one, because this one is important for survival.

One way around this seems to be to hypothesize what it looks like actually happened: somehow, generation after generation certain fear-induced neural net modifications are also somehow structurally the same in the different individuals, and gradually induce corresponding DNA changes to the developmental pathways for the structure of the brain, through some Lamarckian process. But this doesn't seem very satisfactory. I agree that some radical new theory is needed for this, perhaps something along the lines of Sheldrake's ideas, where such acquired characteristics perhaps are transmitted from individuals of the species to some sort of "morphic field".