Guy's main point is correct, but I think he's a bit prescriptive about the number of copies:
For instance if one views a transcript of a parish register one is viewing a second copy of the original.
The information being first recorded in a day book and copied to the register once a week.
Assuming the original contained the correct information still allows for two instances of errors to be made.
A transcript of a Bishop's transcript is a copy of a copy of another copy of the original day book. In other words three instances where errors could occur (assume correct information in the day book.
In some parishes there clearly were day books, because they have survived, but at this distance it's not always possible to tell whether the Bishop's transcript was copied from the register or if that too was from the day book.
Where no day book survives, I don't think we can be certain that there ever was one - and in a small parish with only a handful of CMBs each year I think it's rather unlikely. It's also not certain that the Bishop's transcripts were always copied from the register. In one place I came across I'm pretty sure it was the other way round, with the BT as the first version, written up at the time of the events on loose pages, and the register as a copy of that, since a whole year had been written up very neatly in the same hand and with exactly the same pen and ink etc. (Typically I can't now remember where that was
)
As to the original question of what can be inferred from a discrepancy between the register and the BT, there are a few possibilities, such as:
(a) the extra entry was omitted by mistake in copying from one to the other (or from a day book)
(b) the first version omitted the entry by mistake, but this was realised and corrected in a later one
(c) the version with the extra entry was written up before the event, but for some reason it didn't take place, so the version with the omission is correct
... and probably a few more.
At this distance you can often only make a best guess as to the reason, but looking at the originals is always going to help. There may be clues in the handwriting, or in BTs, a note above the signatures saying what was copied from what.
Arthur
