As I've mentioned before, I'm reading through Richard Bauckham's new tome Jesus and the Eyewitnesses. In the chapter on Eyewitness Memory, Bauckham lists many factors that point to the "reliability of recollective memory" (pp. 330-346). One of those factors is what he calls "Gist and details" (pp. 333-334; 344). Bauckham introduces the concept by saying:

Some writers, particularly those who emphasize the likelihood of inaccuracy in long-term recollective memory, argue that the "gist" of the memory is likely to be accurate, even when the details are not. Barclay maintains that hte recollective memories are "true in the sense of maintaining the integrity and gist of past life events" (Bauckham 333).

As I read this, I flashed back to Bart Ehrman's Gospel of Judas and his presentation on Judas' betrayal. I remembered blogging on the topic and noting how Ehrman focused on the differences of each account of Judas' betrayal and essentially used the differences to cast aspersions on the whole of each account, concluding that the accounts in Matthew and Acts were "impossible to reconcile". My response was to examine each account and determine what was consistent between them and consider that the basis of the event. Using the lingo I just read in Bauckham, I'd say one should examine the different accounts to get at the "gist" of the situation. Bauckham does the same sort of thing when considering Peter's denials of Christ:

A good example of the consistency of the gist along with variation in inessential detail is the story of Peter's three denials of Jesus as told in all four Gospels (Mt 26.58, 69-75; Mk 14.54, 66-72; Lk 22.54-62; Jn 18-15-18, 25-27) (Bauckham 344).

Bauckham goes on to describe the essential and inessential detail of each account and build a likely view of the event based on shared essential detail. The subsection-concluding paragraph wraps it up nicely:

In such examples [here accounts of Peter's denial] we can see that the gist of an eyewitness memory, preserved in all tellings even if other details are not accurately preserved, and the gist of an oral tradition, preserved in all performances even when other details are varied, can readily coincide. This is a most important conclusion for the study of gospel traditions. It is a conclusion that recognizes the realistic extent to which memory can be relied upon, in the case both of the memory of the eyewetness and the memory of the preformer of oral tradition. The transition from the one to the other need not entail a significant decrease in reliability, though of course this is possible (Bauckham 345).