future, or at a distance, we rely on the testimony of others (testing their
reports by considering their credibility as witnesses and the conformity of
the evidence to general experience in like cases); in regard to questions
concerning that which is absolutely beyond experience, _e.g._, higher
orders of spirits, or the ultimate causes of natural phenomena, analogy is
the only help we have. If the witnesses conflict among themselves, or with
the usual course of nature, the grounds _pro_ and _con_ must be carefully
balanced; frequently, however, the degree of probability attained is so
great that our assent is almost equivalent to complete certainty. No
one doubts,--although it is impossible for him to "know,"--that Caesar
conquered Pompey, that gold is ductile in Australia as elsewhere, that iron
will sink to-morrow as well as to-day. Thus opinion supplements the lack of
certain knowledge, and serves as a guide for belief and action, wherever
the general lot of mankind or individual circumstances prevent absolute
certitude.

Although in this twilight region of opinion demonstrative proofs are
replaced merely by an "occasion" for "taking" a given fact or idea "as true
rather than false," yet assent is by no means an act of choice, as the
Cartesians had erroneously maintained, for in knowledge it is determined by
clearly discerned reasons, and in the sphere of opinion, by the balance of
probability. The understanding is free only in combining ideas, not in its
judgment concerning the agreement or the repugnancy of the ideas compared;
it lies within its own power to decide whether it will judge at all, and
what ideas it will compare, but it has no control over the result of the
comparison; it is impossible for it to refuse its assent to a demonstrated
truth or a preponderant probability.

In this recognition of objective and universally valid relations existing
among ideas, which the thinking subject, through comparisons voluntarily
instituted, discovers valid or finds given, but which it can neither alter
nor demur to, Locke abandons empirical ground (cf. p. 155) and approaches
the idealists of the Platonizing type. His inquiry divides into two very
dissimilar parts (a psychological description of the origin of ideas and a
logical determination of the possibility and the extent of knowledge), the
latter of which is, in Locke's opinion, compatible with the former, but
which could never have been developed from it. The rat