Michael Gilleland sends in the following about the scholarly enterprise of hunting for sources (Quellenforschung):

Axioms of Quellenforschung, as outlined by A.E. Douglas, Cicero (Oxford, 1968) = Greece and Rome, New Surveys in the Classics 2, on p. 28:

(i) nobody ever said anything for the first time, particularly if he was a Roman and is extant, (ii) nobody had any general knowledge -- he always had a 'source,' (iii) nobody ever compared sources if he could follow a single one, (iv) nobody ever read a single reputable source if he could use a digest or handbook, (v) lost sources never made mistakes, while extant writers make egregious blunders.

Thanks!