Scientific Progress: not as Straight a Process as it Seems

Posted on 2009-03-19. Permanent link: http://www.rp-photonics.com/spotlight_2009_03_19.html

Author: Dr. Rüdiger Paschotta, RP Photonics Consulting GmbH

Ref.: H. Collins and T. Pinch, “The Golem: What You Should Know about Science”, Cambridge University Press (Cambridge, UK), 2nd edition, 1994; Spotlight article of 2008-07-13

On 2008-07-13, I already reported in the Photonics Spotlight about the not so straight history of the Michelson–Morley experiment. The book referenced above, which I warmly recommend to people interested in the field, gives additional interesting information on that episode (lasting quite a number of decades!), as well as on a number of other cases where scientific progress occurred in a much more complicated way than is usually assumed. Common views on the scientific progress are indeed very over-simplified, and in many cases quite far from an accurate account of reality. In particular, there are often substantial uncertainties about both experimental details and theoretical consequences of physical theories, and it is not easy for scientists to do their work in a totally unbiased way, as it would happen in an ideal world.

I would certainly not go as far as some sociologists, saying that it is not justified for science to claim a higher reliability than other (non-scientific) methods of seeking some understanding of reality. In its essence, the rules of scientific practice are designed to achieve the maximum possible reliability of the knowledge generated, minimizing possible influences from trivial errors, wrong assumptions, personal bias, random influences on experiments, technical faults, etc. I think it is reasonable to assume that this would normally have a profound effect, justifying greater confidence in the results than for many other methods. However, the results of scientific endeavor are certainly not perfectly reliable as (a) the rules are frequently violated (for good or not so good reasons), and (b) even perfect adherence to the rules cannot guarantee absolutely reliable results. Additional errors may be introduced when transferring and interpreting scientific knowledge.

Still, scientific work appears to be clearly the best we can do when seeking objective knowledge. It would seem to be simpler to cultivate friendly personal relations to the creator of the universe, hoping that He will be kind enough to share some of His privileged knowledge. However, the reliable authentication of that dialog is probably even more difficult than scientific work.