icated by a verb,
the substance or concretion of qualities designated by a noun, are
diversities growing up in experience, by no means attributable to the
mere play of sound. The parts of speech are therefore representative.
Their inflection is representative too, since tenses mark important
practical differences in the distribution of the events described, and
cases express the respective roles played by objects in the operation.
"I struck him and he will strike me," renders in linguistic symbols a
marked change in the situation; the variation in phrase is not
rhetorical. Language here, though borrowed no doubt from ancestral
poetry, has left all revery far behind, and has been submerged in the
Life of Reason.

[Sidenote: Yet it vitiates what it represents.]

The medium, however, constantly reasserts itself. An example may be
found in gender, which, clearly representative in a measure, cuts loose
in language from all genuine representation and becomes a feature in
abstract linguistic design, a formal characteristic in expression.
Contrasted sentiments permeate an animal's dealings with his own sex and
with the other; nouns and adjectives represent this contrast by taking
on masculine and feminine forms. The distinction is indeed so important
that wholly different words--man and woman, bull and cow--stand for the
best-known animals of different sex; while adjectives, where declension
is extinct, as in English, often take on a connotation of gender and are
applied to one sex only--as we say a beautiful woman, but hardly a
beautiful man. But gender in language extends much farther than sex, and
even if by some subtle analogy all the masculine and feminine nouns in a
language could be attached to something suggesting sex in the objects
they designate, yet it can hardly be maintained that the elaborate
concordance incident upon that distinction is representative of any felt
quality in the things. So remote an analogy to sex could not assert
itself pervasively. Thus Horace says:

Quis _multa_ gracilis te puer in _rosa_
perfusis liquidis urget odoribus
_grato_, Pyrrha, sub _antro_?

Here we may perceive why the rose was instinctively made feminine, and
we may grant that the bower, though the reason escape us, was somehow
properly masculine; but no one would urge that a _profusion_ of roses
was also intrinsically feminine, or that the _pleasantness_ of a bower
was ever specifically masculine to sense. The epithets _