Seems to me it was David Hume who, perhaps 200 hundred years ago (circa?),
made a telling case for the unreliability of sense data?  Empiricism as
developed through the scientific tradition (which one assumes extends to
the laws of historic evidence) requires a reliance upon the authentication
("documentation") of evidence and the reproducibility of results.
Falsifications do occur but a variety of checks have been built into the
process and the "community" tends to be self-policing...  Scepticism is
probably a necesasary aspect of "the method" and even when in an extreme
form it is applied (by neo-Fascists) to so painful an issue as the Holocaust
(or by neo-primitives to so elemental an issue as the form of the planet)
perhaps it can help us to refine the precision of our methods? (And to
remember that the best response to darkness is more light...?)
Tom Moritz, Calif. Academy of Sciences