Next post will be on 'A better Google Scholar'. Send your best hot takes this way.
Conversation
Replying to
One I find myself returning to often: the "paper" is a pretty poor atomic unit for many of the sense-making activities I want to do. notes.andymatuschak.org/z6X8U64HPkrUaT cc
Yes! Very true! My old unpublished rant on 'coherent credences' to improve science is about claims not papers for this reason. A paper will produce its own data, rely on assumptions from past work etc. Papers are a graph where various things are going on...
1
4
Sometimes bits of the paper are not credible and others are, depending on the technique they used. This would get lost by aggregating everything at the paper level
1