This was a throw-away paragraph in my dissertation, an aside on the difficulties predicting how a source text is used, but in re-reading it, I think I like it very much.
Historians know that they can not normalize historical anomalies and occurrences. History is messy. Things happen with little rhyme or reason, alongside things that take place with long-anticipated intentionality and preparation. An author's use [or creation of a text] is an historical occurrence. Assuming a consistent intentionality [throughout the process] is not grounded in the human experience. Scribes skip lines when copying. Redactors are inconsistent in making changes ("writer's fatigue"). An author, in the process of intentional use of material, is also affected by the size of the page, and a child's interruptions, and the gash on the writing surface, and a sympathy towards a particular character, and the lateness of the hour. Alongside the intentions of the redactor fall these random historical realities that shape the resulting text… Just as it is impossible for an historian to explain the logic and reason behind everything that happens in a period that she is researching, so is it impossible to [fully discern how a text used its sources] from a limited set of presumptuous, consistent criteria. (Weaks, Mark Without Mark, 2010)