onsdag 15 oktober 2014
Science must be without status considerations
One of the most severe fallacies used by pre-publication peer review advocates is to claim that it would be wrong to brainstorm theories just because each single new theory in itself would have a lower chance of being correct than established theories. The fallacy is, of course, to judge methods by the chances of any one theory being correct in the first place. The entire concept of status and prestige must be totally abandoned, and replaced by a pursuit of truth where it does not matter at all who came up with what theory. Questioning must be totally free, as must access to observations and experiment results. Consider, for instance, the fact that it was good for science that Johannes Kepler stole Tycho Brahe's observation notes upon Tycho's death instead of letting Tycho's relatives lock them away. This does not mean that lying about the content of observations and experiments is somehow good. Real science must judge actions by their consequences for science, and not follow associations of "honesty" and "dishonesty" created in an unscientific (statusian) context. The truth is that observations and experiment results are far more timeless than are theories, and therefore collections of observational/experimental data free of theoretical interpretation are very valuable for the progress of science (in sharp contrast to what peer review journals assume when they demand that all papers must contain theoretical interpretations). Any assumptions that pointing out observational/experimental results must somehow be based on a "hidden agenda" just because those results happen to contradict a certain theory must be abandoned as the fallacy it is. Assuming "hidden agendas" behind all observations/experiments that contradict your theory has, for all purposes relevant to scientific testability, the exact same effects as conspiracy theories (even when they are not technically conspiracy theories, the distinctions invariably have zero relevance to scientific testability and therefore does not matter in an objective measure of scientificness).