In physics, theories make predictions of experiments not yet performed.
Herein lies the predicative power of physics. Only in cosmology do
physicists deal with past events. That is why in the past physicists did not
think that cosmology was a science. It is only through the work of Friedmann
and his model of the universe that gave scientific legitimacy to cosmology.
>Most research in other scientific disciplines involves the reconstruction
>of past events. For example, a chemist may perform an experiment in which
>two solutions are mixed. The chemist then analyses the product of the
>reaction and infers, based on chemical theories, what actually took place
>within the test tube. The process itself, the chemical event, was not
>directly observed. Now, whether that analysis was done five minutes after
>the reaction, or a million years after, does not change that fact that a
>past event is being reconstructed from its products or effects.
Physics, and science in general, is based on generalizing experimental data
into laws which are then described by theoretical models. It is the
predictive power of such mathematical models that allow for predications.
Physicists do not reconstruct past events expect when it comes to cosmology.
Past events are the historical prepositions which are generalized into laws.
>Also, evolutionary theories can and are tested using living organisms and
>communities. New insights have been gained from work with bacterial
>cultures, in which evolutionary responses to selective pressures can be
>observed in real time. Evolutionary theories generate testable
>expectations about gene flow and selective pressures within biological
>communities or species populations.
The main issue is if by tinkering with living organisms one can generate a
new organism that is totally different, a difference in kind not degree,
from the original one. If the latter could be accomplished, then that would
lend credence to the theory of common descend.