-evidence,
Inference, and Observation.]

Sect. 76. What is called _formal logic_ is animated with the hope of
extracting these formulations directly from an analysis of the procedure
of thought. The most general logical principles which have appeared in
the historical development of formal logic are _definition_,
_self-evidence_, _inference_, and _observation_. Each of these has been
given special study, and each has given rise to special issues.

_Definition_ has to do with the _formation of concepts_, or determinate
and unequivocal meanings. The universality of such concepts, and their
consequent relation to particular things, was, as we have seen,
investigated at a very early date, and gave rise to the great
realistic-nominalistic controversy.[184:5] A large part of the logical
discussion in the Platonic dialogues is an outgrowth of the earlier
"eristic," a form of disputation in favor with the sophists, and
consisting in the adroit use of ambiguity.[184:6] It is natural that in
its first conscious self-criticism thought should discover the need of
definite terms. The perpetual importance of definition has been largely
due to the great prestige in modern philosophy of the method of
geometry, which was regarded by Descartes and Spinoza as the model for
systems of necessary truth.

_Self-evidence_ is the principle according to which _conviction of truth
follows directly from an understanding of meaning_. In the practice of
his intellectual midwifery, Socrates presupposed that thought is capable
of bringing forth its own certainties. And rationalism has at all times
regarded truth as ultimately accredited by internal marks recognizable
by reason. Such truth arrived at antecedent to acquaintance with
instances is called _a priori_, as distinguished from _a posteriori_
knowledge, or observation after the fact. There can be no principles of
self-evidence, but logicians have always been more or less concerned
with the enumeration of alleged self-evident principles, notably those
of _contradiction_ and _identity_. A philosophical interest in the
mathematical method has led to a logical study of axioms, but with a
view rather to their fruitfulness than their intrinsic truth. Indeed,
the interest in self-evident truth has always been subordinate to the
interest in systematic truth, and the discovery of first principles most
commonly serves to determine the relative priority of definite
concepts, or the correct point of depa