Natural selection is also a problem. Although it can weed out the misfits, natural selection cannot make new things (selection means choosing a few from a greater number). It does not create features but merely selects those that provide a greater survival value, and by doing so only narrows the width of the evolutionary process. Although Neo-Darwinists usually claim that the growth in complexity is the result of adaptation to environment, the appearance of increasingly complex organisms cannot be predicted solely from the work of natural selection upon random mutations.
The classical Darwinian mechanism works mainly to adapt individuals to their existing niches; individual variations do not contribute significantly to the emergence of new, more complex species. Even more importantly, many simple organisms are equally or better adapted to environmental variations than complex ones. Only they can be found in extreme conditions. Some unicellular life forms are spread across different environments much more than complex organisms (with the exception of humans). Evolutionist Gould states that ‘...without question, these earliest and simplest cells, the bacteria and their allies, remain the most abundant, widespread, and successful of all living things' (1988, p.44). If only adaptation directs evolution, evolution should not have moved from one cell organisms. This reasoning can be pushed even further:
If mere survival is the sole desideratum, then it would seem that some rudimentary type of organism would be all that is needed. And there would seem no reason why even a rudimentary type of organism should appear, since it could not hope to rival in longevity the everlasting rocks - but unstable DNA? (Edmunds, 1997, p.159)
Natural selection also cannot adequately explain long term adaptive changes. Some changes have immense consequences, and yet they could not have had adaptive advantages when they happened. One example is bisexual reproduction that increases diversity at great cost. Laszlo points out that ‘such a mechanism, while offering an obvious long-term advantage (the more rapid spread of advantageous mutations) does involve an equally obvious short-term disadvantage (the reduced average number of descendants due to males failing to produce offspring)' (1993, p.169).
Finally, natural selection seems to be based on a circular argument (everything that survives is adaptive and therefore selected, and everything that is selected is adaptive and survives) so it cannot be refuted, which does not make good science. When natural selection is used to explain everything, even mutually contradictory adaptations (e.g. the indistinctive colours of some insects, as well as very distinctive colours of others), in fact, it does not explain much.