And you're misunderstanding the idea here. Say that someone does read a paper which cites a paper that mis-cites. The paper that mis-cites would still be in error and thus the person reading the paper that cites the mis-citing would put faulty information into the paper. The only ways the error wouldn't be propogated is if by a miraculous twist of fate the author who cites the mis-citing paper mis-cites and ends up writing the correct thing (but that's never going to happen), or if the author does an unprecendented amount of work and reads every single paper that is in the bibliography, and then every single paper in the bibliographies of those papers (but that's even more unlikely).
Slashdot Mirror
And you're misunderstanding the idea here. Say that someone does read a paper which cites a paper that mis-cites. The paper that mis-cites would still be in error and thus the person reading the paper that cites the mis-citing would put faulty information into the paper. The only ways the error wouldn't be propogated is if by a miraculous twist of fate the author who cites the mis-citing paper mis-cites and ends up writing the correct thing (but that's never going to happen), or if the author does an unprecendented amount of work and reads every single paper that is in the bibliography, and then every single paper in the bibliographies of those papers (but that's even more unlikely).