peculiar circumstances in the history of a language have arisen, to cause the rapid and general decay and effacement of ancient forms, as in our language and the Romanic, the process of formative composition; though never wholly extinct, has been found too inactive to repair the losses; they have been made up by syntactical collocation, and the language has taken on a prevailingly analytic character.
These considerations and such as these, I am persuaded, furnish a satisfactory explanation of the preponderating tendency to the use of analytic forms exhibited by modern languages; as they also account for the greatly varying degree in which the tendency exhibits itself. But even should they be found insufficient, this would only throw open for a renewed investigation the question respecting the ground of the tendency; the general facts in the history of earliest development of our languages would still remain sure, beyond the reach of cavil, since they are established by evidence which cannot be gainsaid, contained in the structure of the most ancient forms. We are compelled to believe that the formative processes which we see going on, in decreasing abundance, in the historically recorded ages of linguistic life, are continuations and repetitions of the same constructive acts by which has been built up the whole homogeneous structure of inflective speech.
One more theoretic objection to the doctrine of a primitive Indo-European monosyllabism we may take the time to notice, more on account of the respectability of its source than for any cogency which it in itself possesses. M. Renan, namely,[1] asserts that this doctrine is the product of a mistaken habit of mind, taught us by the artificial scholastic methods of philosophizing, and leading us to regard simplicity as, in the order of time, anterior to complexity; while, in fact, the human mind does not begin with analysis; its first acts being, on the contrary, complex, obscure, synthetic, containing all the parts, indistinctly heaped together. To this claim respecting the character of the mental act we may safely yield a hearty assent; but, instead of inferring
- ↑ In his work on the Origin of Language, seventh chapter.
