Tuesday, March 21, 2006

Reading impairs your judgement

Published statements seem to skew the way you interpret your own results. While this seems intuitive the end result is more insiduous. Instead of independent scientific investigations we have a collective "stream of conciousness", which exacerbates the extreems. In other words if we all build on an inital wrong hypothesis we may interpret all our collective results wrongly for a very long time... This doesn't bode well for the idea of wikiscience, "Experiments involving thousands of investigators collaborating on a "paper" will commonplace. The paper is ongoing, and never finished. It becomes a trail of edits and experiments posted in real time — an ever evolving document. "


Bayman said...

On the contary, I think "wikiscience", or the idea of having a more interactive, open publishing environment will help science move forward more rapidly. While holding on to incorrect or more appropriately, incompete, theories or paradigms can slow the pace of science, we cannnot be free of them until better ones are presented to take their place. Thus the more open a network we have for sharing scientific findings, the easier it will be for alternative paradigms to be presented and explored. Of course, by natural selection, only the paradigms that best fit with reality will survive to be integrated into the scientific "stream of consciousness". You could say that wikiscience would increase the reproduction rate of scientific ideas, thus increasing their natural diversity and therefore increasing rate of evolution of science. Of course, selective pressures must remain severe to eliminate the crap.

Anonymous Coward said...

I like the "negative results will have a positive value". Makes me feel better about my stuff... Also at some point they suggest zero-author paper generated by computers. It's already hard to keep up with the litterature, what happens if nobody reads it? Is it still knowledge? Can you hear the tree falling in the forest. To address your point, the problem with consensus-building is that you reduce idea variability and have less chance to find better ones. It's like going full steam but after only one idea at a time. Just think how hard it is to publish stuff that contradicts dogma. Now what if there was only one BIG paper?

Bayman said...

Intersting... I guess there's a bit of a paradox - even random computer-generated data wouldn't be so bad, as long as you use human critical reasoning to fit things together and eliminate the garbage. Of course relying on human critical reasoning (ie "peer review") then allows consensus building to eventually creep in....Although this system is imperfect by necessity, history has shown that it works, and that over time imperfect theories get better and the wrong ones knocked over, improved paradigms eventually become impossible to ignore when the facts come to light.