another poor creature convention.' But the poor creature convention in the
end proves too much for all the rest: for we do not ask what is the origin
of words or whether they are formed according to a correct analogy, but
what is the usage of them; and we are compelled to admit with Hermogenes in
Plato and with Horace that usage is the ruling principle, 'quem penes
arbitrium est, et jus et norma loquendi.'

(8) There are two ways in which a language may attain permanence or fixity.
First, it may have been embodied in poems or hymns or laws, which may be
repeated for hundreds, perhaps for thousands of years with a religious
accuracy, so that to the priests or rhapsodists of a nation the whole or
the greater part of a language is literally preserved; secondly, it may be
written down and in a written form distributed more or less widely among
the whole nation. In either case the language which is familiarly spoken
may have grown up wholly or in a great measure independently of them. (1)
The first of these processes has been sometimes attended by the result that
the sound of the words has been carefully preserved and that the meaning of
them has either perished wholly, or is only doubtfully recovered by the
efforts of modern philology. The verses have been repeated as a chant or
part of a ritual, but they have had no relation to ordinary life or speech.
(2) The invention of writing again is commonly attributed to a particular
epoch, and we are apt to think that such an inestimable gift would have
immediately been diffused over a whole country. But it may have taken a
long time to perfect the art of writing, and another long period may have
elapsed before it came into common use. Its influence on language has been
increased ten, twenty or one hundred fold by the invention of printing.

Before the growth of poetry or the invention of writing, languages were
only dialects. So they continued to be in parts of the country in which
writing was not used or in which there was no diffusion of literature. In