In the early days of text mining and search engines in France there was the heady idea that once search was perfected we could answer most of the questions of science just running through the databanks Things like googlewhacks seemed to hint at a proof of concept There was the whole question of exactly how smart we could become if we had access to a search engine But a lot of prosaic obstacles arose like Nickwords (the opposite of keywords) or trust in source, and the sheer obscurity of Scientific English
Clemens,
I think I could get a short chapter on that theme out