Cosma Shalizi's "null model of scientific inquiry" [1] seems relevant here. Basically it's a model where all published findings are due to the file drawer effect [2]. In such a model, scientific results will indeed have a "half-life" until they are overturned.<p>[1] <a href="http://masi.cscs.lsa.umich.edu/~crshalizi/weblog/698.html" rel="nofollow">http://masi.cscs.lsa.umich.edu/~crshalizi/weblog/698.html</a><p>[2] <a href="http://en.wikipedia.org/wiki/Publication_bias" rel="nofollow">http://en.wikipedia.org/wiki/Publication_bias</a>
"IN PRIMARY school Babbage learned that there were nine planets in the solar system. None were known to exist outside it."<p>There were only 7 planets known when Charles Babbage was a youth. Neptune was discovered when he was an old man, and Pluto was not discovered until almost 60 years after his death.
There is a excerpt from chapter 1 of the book at the authors website.<p><a href="http://arbesman.net/the-half-life-of-facts/" rel="nofollow">http://arbesman.net/the-half-life-of-facts/</a>