Posted: Jun 12, 2010 11:20 am
by Shrunk
TMB wrote: This misses two points.

The first is that nothing is exactly repeatable in every way, meaning that there are always factors of unreliable in the process. Even science method cannot ensure that everything can be exactly as it was. This means that science itself is subject to its own critique of method. This means that regardless of how good the fidelity of repeat is, it should always be viewed as possibly flawed.

Right. Hence the need to repeatedly try to falsify a hypothesis even once it has seemingly been confirmed experimentally. Technically you're correct; no theory can be said to be correct with 100% certainty, as you never know if the next observation will falsify it. However, after a sufficiently large number of falsification attempts have failed, it can be said for all practical purposes that the theory is correct with a certainty that approaches 100%.

Which is precisely the error Nancy Malik continually makes in this thread: Accepting isolated, uncorroborated observations as evidence even when they have been refuted by further, more systematic and reliable observations.

The second point exists because of the first. As with the evidence based legal systems, we are simply not in a position to repeat the past with exact fidelity. This means that real stuff that did happen, cannot be proven, although logically we know it is not because the past did not happen in a certain way, just that we are unable to effectivly repeat history. This means that we are forced by the method to dscount things that do not pass this test. This does not mean these things exist or are effective, just that SM is not effective in assessing them.

Right. So those things go into the "unproven, further evidence needed" or "unprovable, we'll never know" categories.

Homeopathy, OTOH, clearly belongs in the "disproven, don't bother with it anymore" category.