ments, methodical doubt
becomes _methodical distrust_.

The historian ought to distrust _a priori_ every statement of an author,
for he cannot be sure that it is not mendacious or mistaken. At the best
it affords a presumption. For the historian to adopt it and affirm it
afresh on his own account implies that he regards it as a scientific
truth. To take this decisive step is what he has no right to do without
good reasons. But the human mind is so constituted that this step is
often taken unconsciously (cf. book ii. chap. i.). Against this
dangerous tendency criticism has only one means of defence. We must not
postpone doubt till it is forced upon us by conflicting statements in
documents; we must _begin_ by doubting. We must never forget the
interval which separates a statement made by any author whatsoever from
a scientifically established truth, so that we may continually keep in
mind the responsibility which we assume when we reproduce a statement.

Even after we have accepted the principle and resolved to apply this
unnatural distrust in practice, we tend instinctively to free ourselves
from it as soon as possible. The natural impulse is to perform the
criticism of the whole of an author, or at least of the whole of a
document, in the lump; to divide authorities into two categories, the
sheep on the right, the goats on the left; on the one side trustworthy
authors and good documents, on the other suspected authors and bad
documents. Having thus exhausted our powers of distrust, we proceed to
reproduce without discussion all the statements contained in the "good
document." We consent to distrust suspected authors such as Suidas or
Aimo, but we affirm as established truth everything that has been said
by Thucydides or Gregory of Tours.[145] We apply to authors that
judicial procedure which divides witnesses into admissible and
inadmissible: having once accepted a witness, we feel ourselves bound to
admit all his testimony; we dare not doubt any of his statements without
a special reason. Instinctively we take sides with the author on whom we
have bestowed our approval, and we go so far as to say, as in the law
courts, that the burden of proof rests with those who reject valid
testimony.[146]

The confusion is still further increased by the use of the word
_authentic_, borrowed from judicial language. It has reference to the
origin only, not to the contents; to say that a document is authentic is
merely to say that