in the
scientific treatment of human speech, than the dialect of the Hottentots.

We saw, secondly, that after the first practical acquisition and careful
analysis of the facts and forms of any language, the next and most
important step is the classification of all the varieties of human speech,
and that only after this has been accomplished would it be safe to venture
on the great questions which underlie all physical research, the questions
as to the what, the whence, and the why of language.

We saw, thirdly, that there is a distinction between what is called
history and growth. We determined the true meaning of growth, as applied
to language, and perceived how it was independent of the caprice of man,
and governed by laws that could be discovered by careful observation, and
be traced back in the end to higher laws, which govern the organs both of
human thought, and of the human voice. Though admitting that the science
of language was more intimately connected than any other physical science
with what is called the political history of man, we found that, strictly
speaking, our science might well dispense with this auxiliary, and that
languages can be analyzed and classified on their own evidence
particularly on the strength of their grammatical articulation, without
any reference to the individuals, families, clans, tribes, nations, or
races by whom they are or have been spoken.

In the course of these considerations, we had to lay down two axioms, to
which we shall frequently have to appeal in the progress of our
investigations. The first declares grammar to be the most essential
element, and therefore the ground of classification in all languages which
have produced a definite grammatical articulation; the second denies the
possibility of a mixed language.

These two axioms are, in reality, but one, as we shall see when we examine
them more closely. There is hardly a language which in one sense may not
be called a mixed language. No nation or tribe was ever so completely
isolated as not to admit the importation of a certain number of foreign
words. In some instances these imported words have changed the whole
native aspect of the language, and have even acquired a majority over the
native element. Turkish is a Turanian dialect; its grammar is purely
Tataric or Turanian. The Turks, however, possessed but a small literature
and narrow civilization before they were converted to Mohammedanism. Now,
the language of