feonixrift: (Default)
[personal profile] feonixrift
[Error: unknown template qotd]

Normally I don't do these, but this hit my "someone is wrong on the internet" button.

First, it depends on whether we're talking about science as the conceptual structure for knowledge acquisition, science as it's actually practiced, science as it's claimed that it ought to be practiced or what's commonly thought of as science.

Then, it depends on whether the question is actually about explanations or proofs. Since this is related to science, we ought at least try to be precise.

Science (the method) does not prove anything, ever. Nor, in fact, does it explain anything. It predicts things, it is forced to change when its predictions are wrong, over time hopefully its predictions become more accurate. The scope about which it can predict and still be 'scientific' only includes predictions the validity of which could be tested. Not everything is testable, therefore not everything is within its potential predictive scope.

But even what a theory correctly predicts, it does not precisely explain, because there is no proof that the reasons contained within the theory are the reason why that event occurred in keeping with it. (cf. the history of the concept of epicycles in astronomy) Clinging to explanations rather than accuracy of prediction does not improve progress in scientific theory.

Science (as practiced) proves nothing, but provides potential explanations for a great many things. Solely potential, mind, they are not necessarily accurate. Thus, unproven, and unprovable. But it reaches well beyond the realms of direct experimental testing, even as far as the formation of galaxies. (Beyond that, well, I'm pretty sure the physicists have digressed into philosophy, but it remains true that they do so under the title of science.)

Science (as supposedly ought to be) would follow its method exactly. But frankly, this doesn't work. As with Mathematics, the course into the æther of the wild hunch must be followed first, then the formalities connecting it back to previous knowledge may be constructed. Unlike with Mathematics, those connections are not in the form of proof, but simply of evidence. Predictions and test cases, nothing more. Subject, due to messy environments, to the tedium of statistics before they can be declared real.

Science (as commonly seen) is expected to explain a lot, and prove itself. Fortunately, only to the common level of proof, which allows that the apple always falls down. It is put on a pedestal, as a source of knowledge, expected to provide sureties. At which, from a layman standpoint, it's quite effective. Until you try to actually talk to a scientist. That outward certainty which it is valid to a first approximation to trust, stands upon the uncertainty of those creating it. They must be continually aware that none of it is set in stone, for it to become firm enough that everyone else can treat it as set in stone.

But this is easily latched upon to say that their uncertainty, as experts, renders something surely uncertain. Or the opposite, when they do become certain and ignore evidence, toppling a piece of the tower with them until others can rebuild it in its proper shape, how untrustworthy they must seem then. So what of the educatedly uncertain layman? Are they more accommodating of the necessity that scientists remain, as a whole uncertain? Or fated by the curiosity that accepting uncertainty engenders, to become scientists themselves, if only in some small measure?

No, we do not explain. We do not prove. We conjecture, explore and draw maps. The map is not the territory, but it just might get you there.

Date: 2009-08-14 02:27 pm (UTC)
From: [identity profile] dancinglights.livejournal.com
oh this is lovely. I wanted to natter in a similar way about this one, but couldn't quite find the perfect words. These are very lovely words, and they explain the conundrum well.

Date: 2009-08-14 02:55 pm (UTC)
From: [identity profile] feonixrift.livejournal.com
Thank you! I meandered a bit, I'm glad it worked.

Expand Cut Tags

No cut tags


feonixrift: (Default)

February 2012

   1 234

Style Credit

Page generated Sep. 21st, 2017 01:30 am
Powered by Dreamwidth Studios