179. Linguistic Diversification
Language tells a similar story. The American Indian languages certainly appear to be diverse. It has been customary to reckon about a hundred and fifty distinct speech families in North and South America (§ [50]). But many of these are imperfectly known; of late several Americanist philologists have been inclined to see definite resemblances between numerous tongues that are superficially different. Buried and disguised resemblances are being noted, which point to original unity. Thus the number of genetically distinct families or language stocks is shrinking. The number to be ultimately recognized bids fair to be small.
Old World conditions are at bottom more similar than at first glance seems. English and the modern Hindu languages, such as Bengali, although certainly related, are quite different from each other. The proof of their common Indo-European origin rests largely on the similarities between their ancestral forms, Anglo-Saxon and Sanskrit. These in turn are tied together more closely by the connecting evidence of other ancient languages, such as Greek and Latin. Take away these extinct tongues and the modern transitional ones, imagine English and Bengali to be the only representatives of Indo-European, and it is doubtful whether their common parentage could be wholly proved. The relationship would certainly not be readily recognizable; the most painstaking analysis would reveal so many words wholly peculiar to each language, and so many exceptions to every suggestion of regular sound equivalences, that conservative philologists would perhaps refuse to commit themselves on the problem of a single origin.
This imaginary situation parallels the actual one in American linguistics. Not a single ancient form of speech has been preserved. Many living ones are inadequately known. The fact that some enthusiast has compiled a grammar of Nahuatl or Quechua or Eskimo does not necessarily mean that he has dissected out its whole structure. A book devoted to a language may be as superficial as it looks learned. And the man who really knew Nahuatl has usually concerned himself very little with Quechua. So far as he might become acquainted with it, it would appear so different that the pressing of comparisons would seem sterile. Thus the great diversity of American languages came to be accepted not because any one believed it to have been really established, but because until recently no critical scholar considered himself able to establish serious connections. It has been a case of unproved rather than disproved unity of origin.
If the Indo-European languages were not our own but those of a strange race and therefore known to us much less intensively; if the history of their ancient forms were obliterated instead of preserved for us for over three thousand years; and if they were allowed a period of ten thousand years in which to have differentiated, philology would probably be assigning them to several distinct stocks. Multiply by three the amount of difference which Bengali shows from Sanskrit, and by six that of English from Anglo-Saxon, and a degree of divergence might be attained roughly comparable to that between Nahuatl and Quechua, or Maya and Iroquois. This is not an assertion that Nahuatl and Quechua are related. It is a claim that in the light of present knowledge they might have been one language ten thousand years ago. A single people with a single speech could well have given rise in so long a period as that—three hundred generations—to languages that now seem so different.
And at that, there is no reason for believing that all the American languages are necessarily derivable from a single mother tongue. There might have been half-a-dozen or a dozen idioms in use among as many populations which moved out of Asia into Alaska. For of course it is improbable that the migration was an isolated, unitary event. More likely it filled a period of some length, during which a succession of waves of population lapped from one continent into the other. Each of these waves, which only the perspective of ages has merged into the appearance of a single movement, may have brought its own speech, from which in time there branched out languages that ultimately became so differentiated as to appear now like distinct families. Not that it is known that this happened; but it seems inherently plausible that it might have happened, and there is no present evidence to the contrary.
In short, philology interposes no obstacle to the acceptance of the date which has been assumed as roughly defining the period of the peopling of America.