ddle Ages were emblematic
devices; but the architecture of the cathedrals was largely symbolic.
Both agree in aiming to aid the imagination and the memory, and both may
appeal to any special sense, although the majority are addressed to
sight alone.

Symbolism has not received the scientific treatment which has been so
liberally bestowed on mythology. The first writer who approached it in
the proper spirit was Professor Creuzer.[200-1] Previous to his labors
the distinction between pictographic and symbolic art was not well
defined. He drew the line sharply, and illustrated it abundantly; but he
did not preserve so clearly the relations of the symbol and the myth.
Indeed, he regarded the latter as a symbol, a "phonetic" one, to be
treated by the same processes of analysis. Herein later students have
not consented to follow him. The contrast between these two expressions
of the religious sentiment becomes apparent when we examine their
psychological origin. This Professor Creuzer did not include in his
researches, nor is it dwelt upon at any length in the more recent works
on the subject.[201-1] The neglect to do this has given rise to an
arbitrariness in the interpretation of many symbols, which has often
obscured their position in religious history.

What these principles are I shall endeavor to indicate; and first of the
laws of the origin of symbols, the rules which guided the early
intellect in choosing from the vast number of objects appealing to sense
those fit to shadow forth the supernatural.

It may safely be assumed that this was not done capriciously, as the
modern parvenue makes for himself a heraldric device. The simple and
devout intellect of the primitive man imagined a real connection between
the god and the symbol. Were this questioned, yet the wonderful
unanimity with which the same natural objects, the serpent, the bird,
the tree, for example, were everywhere chosen, proves that their
selection was not the work of chance. The constant preference of these
objects points conclusively to some strong and frequent connection of
their images with mythical concepts.

The question of the origin of symbols therefore resolves itself into one
of the association of ideas, and we start from sure ground in applying
to their interpretation the established canons of association. These, as
I have elsewhere said, are those of contiguity and similarity, the
former producing association by the closeness of succession