A certain number of languages may remain isolated without this fact affording any evidence against the specific unity of man. In all philological schools it is acknowledged that languages are variable and perishable. Now we do not know all the dead languages, and if some of the links in the chain are wanting it will at once be evident that relations which formerly existed have been lost to us for ever.
Let anyone, moreover, refer to the observations of Lubbock upon roots, and he will at once admit that a certain number among them can scarcely be common to all languages. Those who hold that language is not of divine origin, but a human invention and creation, cannot help adopting the conclusions of the learned Englishman on this point. Now, however few these radical differences may be, they necessarily involve irreducibility, which cannot, however, on that account be invoked as an argument against monogenism.
In support of this conclusion, I am fortunate enough to be able to appeal to the testimony of a judge, both competent and trustworthy. Whitney, in his work upon “The Life of Language,” has examined the same question. With Crawfurd and M. Hovelacque, the American linguist admits that there are linguistic families which cannot be referred to a common origin. He does not, however, stop at the bare fact; he demonstrates and discusses the causes of it. He then gives, in the following terms, the general conclusion of this discussion: “The incompetency of the science of philology to decide upon the unity or diversity of human races appears to be completely and irrevocably demonstrated.”
However this may be, the results thus acquired bring to light a fact, the importance of which ought not, it seems to me, to be overlooked. Taking as guide the work of a man whose competency is above dispute, arranging the tables of the linguistic families admitted by M. Maury, and representing by lines the relations pointed out by this learned writer, we see that there exists between one language and another an intercrossing of characters extremely analogous to that which I have so often pointed out in human groups. No one has supported the hypothesis of the multiple origins of languages more resolutely than Agassiz. In the memoir, which I attacked from a geographical point of view, he expressed himself very clearly upon this point. Since then he has developed the same ideas. I have already said that, in his opinion, mankind was created by nations, that each received, with its physical features, its particular language, developed in every direction, and just as characteristic as the voice of an animal species. I feel it necessary to insist upon this point here, and to quote the text itself: “Let anyone follow upon a map,” says Agassiz, “the geographical distribution of the bear, the felidæ, the ruminants, the gallinaceæ, or of any other family: we can prove, with just as much evidence as any philological research can for human languages, that the growling of the bear of Kamschatka is allied to that of the bear of Thibet, of the East Indies, of the Sonda Islands, of Nepaul, Syria, Europe, Siberia, the United States, the Rocky Mountains, and the Andes. Yet all these bears are considered to be distinct species, having in no way inherited voice from each other. Nor have the different human races done so. All this is equally true of the crowing of the gallinaceæ, of the quacking of ducks, as well as of the song of thrushes, who all pour forth their gay and harmonious notes, each in their own dialect, which is neither inherited nor derived from another, although all sing in thrush language. Let philologists study these facts, and if they are not absolutely blind to the signification of analogies in nature, they will themselves come to doubt the possibility of placing any confidence in philological arguments employed to prove genetic derivation.”
Agassiz is logical, and he exhausts the consequences of his theory. But he forgets one important fact which may be opposed to all those who, either fully or partially, embrace this order of ideas.
No animal species has ever changed its voice for that of a species nearly allied to it. An ass’s colt, reared by a mare and isolated in the midst of horses, never forgets its bray or learns to neigh. While, on the contrary, it is well known, that a White, if placed in earliest infancy in the midst of Chinese or Australians, will only speak their language. The converse is equally true.
The reason of this is that the animal voice is a fundamental character, adhering evidently to the nature of the being, susceptible of slight modification, but incapable of disappearing, or of transference as a whole; it is a specific character.
Human language is entirely different. It is essentially variable, and subject to modification from one generation to another it is subject to transformation; it borrows and loses; it may be replaced by another; it is evidently subordinate to the intelligence and to the conditions of life. We can only, therefore, regard it as a secondary character; a character of race.
From the linguistic point of view, the specific attribute of man is not the special language which he employs, it is the faculty of articulation, speech, which has given him the power of creating a primitive language, and to vary it infinitely by means of his intelligence and will, more or less influenced by innumerable circumstances.
Here, again, I am fortunate enough to be able to support opinions, which I have long maintained, by the conclusions of Whitney upon this point. “Now,” says this learned linguist, “to pretend, in order to explain the variety of languages, that the power of expression has been virtually different in different races, that one language has contained, from its origin and in its primitive materials, a formative principle which is not in others; that the elements employed for a formal usage were formal by nature, and so on,—all this is pure mythology.”