Sanscrit and Greek. They hardly enable us to approach any nearer the
secret of the origin of language, which, like some of the other great
secrets of nature,--the origin of birth and death, or of animal life,--
remains inviolable. That problem is indissolubly bound up with the origin
of man; and if we ever know more of the one, we may expect to know more of
the other. (Compare W. Humboldt, 'Ueber die Verschiedenheit des
menschlichen Sprachbaues;' M. Muller, 'Lectures on the Science of
Language;' Steinthal, 'Einleitung in die Psychologie und
Sprachwissenschaft.'

...

It is more than sixteen years since the preceding remarks were written,
which with a few alterations have now been reprinted. During the interval
the progress of philology has been very great. More languages have been
compared; the inner structure of language has been laid bare; the relations
of sounds have been more accurately discriminated; the manner in which
dialects affect or are affected by the literary or principal form of a
language is better understood. Many merely verbal questions have been
eliminated; the remains of the old traditional methods have died away. The
study has passed from the metaphysical into an historical stage. Grammar
is no longer confused with language, nor the anatomy of words and sentences
with their life and use. Figures of speech, by which the vagueness of
theories is often concealed, have been stripped off; and we see language
more as it truly was. The immensity of the subject is gradually revealed
to us, and the reign of law becomes apparent. Yet the law is but partially
seen; the traces of it are often lost in the distance. For languages have
a natural but not a perfect growth; like other creations of nature into
which the will of man enters, they are full of what we term accident and
irregularity. And the difficulties of the subject become not less, but