Why would anyone care what, say, Charles Babbage was trying to do, when he didn't actually do it in any sort of useful way? Surely this is largely how knowledge is produced—over time, with lots of false starts, and language and technique that is not yet adequate to the problem at hand.