This question isn’t about atheism per se, but it is certainly relevant since most atheists seem to prefer science to religion as a system for understanding the world.
Science is based on a few assumptions: 1) basic inductive reasoning. If a pattern seems to repeat enough times then it’s safe to assume that it will continue more or less identically forever. While doubting this assumption means doubting whether the sun will rise in the morning, it’s still unprovable.
2) Once inductions are made and accepted as natural laws (I understand that science regards such “laws” as hypotheses, so don’t get sidetracked), we assume that the natural laws (such as gravity, quantum physics etc, obviously certain natural laws are location specific) we observe in the local universe will be the same everywhere in space and time, even during events unwitnessed by any recorded person. Doubting this assumption can mean that gravity may reverse somewhere in the far, unobserved universe, or that some 4000 years ago the universe was created out of nothing. We have no evidence that either of these statements is false. Carbon dating and cosmic microwave radiation don’t prove that God didn’t create a world to appear billions of years old (creationists are so stupid! why do they argue against the science rather than allowing God to be complex and wonderful?).
3) perhaps most importantly, science values certain kinds of apprehension over others. Science does not value private gnosis, science does not value learning that comes from within (aside from its psychological value). I can’t really see why. Science values testability but it doesn’t accept evidence that can’t be demonstrated physically. Is there an a priori defense for this position?
Without these assumptions, science is nonsense, and yet these assumptions can never be proven or disproven. This is inconsistent with the value that most atheists give to falsifiability and the validity of science that derives from falsifiability. The third point raises the question of what constitutes falsifiability, what constitutes an acceptable test for a theory? I think these questions raise serious doubts about claims that atheists are using 'reason’ and not faith to judge the world.
Posted: June 19th 2007