I would like to believe that two texts that cite the same third are related through this citation, a bit like two people who grew up with no knowledge of one another may be delighted to discover that they share a parent. This genealogy would thus be the occasion to trace resemblances between the two, again much like one does with siblings.
Of course, while any given author may be presumed to know whom and what she cites, there is no need to assume any knowledge of all or any of the other texts that cite the same source. The family relation between them would thus be a hidden one and any resemblances purely accidental.
Comparative literature is defined by comparison, yet there are no rules for what constitutes one. Presumably, the self-same cannot be compared to itself and thus comparative literature requires a minimal amount of difference that it then can proceed to cross. Alternatively, it could uncover differences within the apparently same.
This has two implications for comparing two texts on the basis of them citing the same third:
Firstly, the shared citation provides the point of similarity around which their differences crystallise.
Secondly, returning to the cited text after having read the texts citing it, one may expect to find differences within this self-same text and thus prove the already granted point that no one text can be read twice, let alone by different readers.
Thoughts fallen by the wayside 