Page:Language and the Study of Language.djvu/418

396 history can furnish, and all that can be derived from archæology, to correct and confirm its conclusions. So intricate and difficult of solution is the problem set before us in the beginnings of history, the origin and ultimate connections of races, that, as we have good reason to fear, our utmost efforts, our most cunning combinations of all attainable evidence, from whatever sources derived, will never bring us to a distinct and confident answer. For a little way, history and tradition are our chief guides; then, the study of language conducts us somewhat farther, although with feebler and more uncertain steps; while physical science claims to give us a few glimpses, we know not yet of what reach or sweep, into a still remoter past. And as, in investigations of this trying character, it is of no small consequence to know what are the limits and defects of the evidence with which we are dealing, that we may not waste our strength, and prepare for ourselves bitter disappointment, by searching for conclusions where none can possibly be found, we entered upon an inquiry as to whether it was within the province of linguistic science to determine the vexed question of the unity or multiplicity of the human race; and we found that this was not the case. The beginnings of language, in at least a part of the recognized families of languages, are too much covered up and hidden under the products of later growth for our eyes ever to distinguish them with any even tolerable approach to certainty; and the correspondences which have been already, or may be hereafter, pointed out between the linguistic material of different languages, now reckoned as belonging to diverse families, may be so plausibly explained as the effects of chance that they can never be accepted as the sure result and sign of a common linguistic tradition. Our conclusion here was, that human languages might well have become as different as we now find them to be, even though all of them descended from the rudimentary and undeveloped dialect of some single original family or tribe; while, on the other hand, considering the acknowledged unity in diversity of human nature, we should not expect to find languages any more unlike than they actually are, if there had been a separate Adam and Eve for each one of a dozen or more human races.