Reconsidering the logical structure of the theory of natural selection

Natural selection has been criticized as a tautology. This would be a major problem for evolutionary biology, if true, because tautological statements can't be falsified and, therefore, can't be scientific. There is merit to this critique insofar as the theory of natural selection is indeed generally described in a tautological manner. However, natural selection can be described non-tautologically if we're careful. Natural selection should be defined as the theory that attempts to predict and retrodict evolutionary change through environmental forces acting upon organisms. However, this re-framing comes at a cost: it reveals, based on our current knowledge of evolutionary forces, the lack of ability to make accurate predictions about expected changes except in the most simple of circumstances. I suggest that evolutionary biologists should work to develop "principles of evolution," based on an expanded and more focused research program designed to identify common trends in evolution, which will ultimately allow us to make more accurate predictions and retrodictions about evolution.