to think of the masculine of vacca, vache, Kuh; the opposite of fortis, fort, stark; the Latin, French, and German ways of expressing "to make big" and "to make small." The issue is hardly doubtful.
Again, the languages upon whose vocabulary and grammar the international language is to be based must be Aryan (Indo-European). This is a practical point. The non-European peoples will consent to learn "simplified Aryan" just as they are adopting Aryan civilization; but the converse is not true. The Europeans will go without an international language rather than learn one based to some extent upon Japanese or Mongolian. The only prescription for securing a large field is—greatest ease for greatest number, with a handicap in favour of Europeans, to induce them to enter.
PART III
THE CLAIMS OF ESPERANTO TO BE TAKEN SERIOUSLY:
CONSIDERATIONS BASED ON THE STRUCTURE OF THE LANGUAGE ITSELF
I
esperanto is scientifically constructed,
and fulfils the natural tendency in evolution of language
All national languages are full of redundant and overlapping grammatical devices for expressing what could be equally well expressed by a single uniform device. They bristle with irregularities and exceptions. Their forms and phrases are largely the result of chance and partial survival, arbitrary usage, and false analogy. It is obvious that a perfectly regular artificial language is far easier to learn. But the point to be insisted on here is, that artificial simplification of language is no fantastic craze, but merely a perfect realization of a natural tendency, which the history of language shows to exist.
At first sight this may seem to conflict with what was said in [Part I., chap. x]. But there is no real inconsistency. As pointed out there, there is no reason to think that Nature, left to herself, would ever produce a universal language, or that a simpler language would win, in a struggle with more complex ones, on account of its simplicity. But this does not prevent there being a real natural tendency to simplification—though in natural languages this tendency is constantly thwarted, and can never produce its full effect.
How, then, is this tendency to simplification shown in the