The Theory of a Common Origin for All Languages
OF all the great changes in thought which the present century has witnessed, perhaps none is more striking than that which has occurred in our methods of studying the beginnings of human culture. The discoveries of Grimm and Bopp in comparative philology, the decipherment of mysterious inscriptions in Egypt and Assyria, tho study of legal archæology illustrated by Sir Henry Maine, the doctrine of survivals so ably expounded by Mr. Tylor, and especially the geologic proof of the enormous antiquity of tho human race, together with the wide-reaching and powerful speculations of Mr. Darwin, have all contributed to bring about this change. So completely has our point of view been shifted by these various theories and discoveries that many speculations which at the beginning of the present
century possessed an absorbing interest have now come to seem frivolous or irrelevant; and nothing can better illustrate the extent of the change than the fate of some of these speculations. It is not many years since ethnologists were racking their brains to show how the North American Indians might have come over from Asia; and there was felt to be a sort of speculative necessity for discovering points of resemblance between American languages, myths, and social observances and those of the Oriental world. Now the aborigines of this continent were made out to be Kamtchatkans, and now Chinamen, and again they were shown, with quaint erudition, to be remnants of the ten tribes of Israel. Perhaps none of these theories have been exactly disproved, but they have all been superseded, and have lost their interest. We now know that in the earliest post-Pleiocene times, at least a hundred thousand, and probably several hundred thousand, years ago the American continent was inhabited by human beings. The primeval Californian skull, moreover, resembles the modern American Indian type, and is not to be confounded with Old World skulls. It is probable, therefore, that far back in post-Pleiocene times, before the great glacial period, the ancestors of the American Indians had already become distinguished from the races of Asia. In these remote ages the two continents may very likely have been joined together at their northeastern and northwestern extremities. At any rate, whatever view we may ultimately adopt, we feel that all theories of the recent colonization of America by Kamtchatkans, or Chinamen, or the ten tribes of Israel are superseded and laid on the shelf. That recent migrations may have occurred is quite another affair. Theories like those of Brasseur de Bourbourg are still to be treated on their own merits, independently of general considerations. But one now perceives, in reading them, that they were dictated by a kind of speculative necessity which we no longer feel, because our whole point of view has been shifted.
In similar wise have fared the innumerable plans which formerly occupied the attention of scholars for colonizing the whole world from the highlands of Armenia. The ethnological information contained in the book of Genesis is of great interest and value, but so far from relating to the whole human race, it totally ignores the larger part of the world, and is concerned only with the peoples of which an inhabitant of Syria might be expected to know something. Long before any possible date for the diffusion from Armenia there described, we know that populous and stationary communities flourished on the banks of the Nile and the Euphrates; while savage or barbarous tribes, using stone hatchets and Hint-headed arrows, wandered through the primeval forests of Europe and America. Armenia retains its interest, to some extent, as a possible starting-point, but only in connection with the Semitic race and its neighbors, — so thoroughly have our notions been remodeled.
Old-fashioned speculations concerning the primitive unity of human speech have similarly fallen into discredit. Previous to the detection of the kinship between the various forms of Aryan speech, no end of books were written to prove that all known languages were in some way descended from Hebrew; not that there was any warrant for such an opinion, either in Scripture or in the general probabilities of the case, but that the preëminence of Hebrew as the language of Jehovah’s chosen people and the vehicle of divine revelation, created a speculative need for proving it to be the original uncorrupted dialect of mankind. Since the establishment of the Aryan family of languages, it. has still been felt necessary to prove that all existing varieties of speech have had a common origin, and as a step toward this end great learning and ingenuity have been expended in the attempt to detect some primordial similarity between the Semitic languages and languages of Aryan descent.
It is not too much to say that all this learning and ingenuity have been utterly wasted. Apart from a few casual coincidences, as in the Hebrew and Sanskrit word’s for six, there is not a trace of similarity between the Semitic and the Aryan vocabularies ; while as regards both inflection and syntax, the entire structure of these two families of speech is so radically unlike, that only the most desperate feeling of speculative necessity could ever have induced any one to seek a common original for the two. But after getting irretrievably worsted in the encounter with facts, this speculative craving is now outgrown and laid aside with the others. The antiquity of the human race again comes in to alter entirely our stand-point. Considering how multifariously language varies from age to age, and considering that mankind has doubtless possessed the power of articulate speech for some thousands of centuries, it no longer seems worth while to seek immediate conclusions about primitive speech from linguistic records which do not carry us back more than four or five thousand years.
From the vantage-ground which we now occupy, it is not difficult to see that the hypothesis of a single primeval language, from which all existing languages have descended, involves an absurd assumption. Those who maintain such an hypothesis, in so far as their statements have any definite and tangible meaning, must mean that all existing languages stand in relation to the hypothetical primitive language very much as French and Italian stand in relation to Latin, or English and German to Old Teutonic, or Latin and Old Teutonic to Old Aryan. But in point of fact the case is very different from this. We know that French and Italian are differently modified forms of Latin, because we can trace the modern words directly back to their ancient prototypes, and verify by the aid of written documents their various changes of form and meaning. After carrying on for a while this process of comparison, we find that the modern words vary from the ancient according to certain welldefined rules, which are different for French and Italian, but are singularly uniform for each language. So unmistakable is the regularity of the system of changes, that if all record of Latin were to be swept away we might still reconstruct the language from a comparative study of its modern descendants. Mois and mese, for example, the French and Italian words for “ month,” would give us the Latin mensis, and nothing else ; and so on throughout. In
similar wise, although the Old Aryan language has left no written documents to tell us of its grammar and vocabulary, we have nevertheless detected such a regular system of phonetic changes among the languages which have descended from it that we have been already enabled to go very far toward reconstructing this extinct tongue. Month and mensis, for example, carry us back, with little less than absolute certainty, to an Old Aryan mansa ; and so on as before, though here the inquiry is an abstruse one, requiring patience and sound judgment, and there is room enough for doubt in many cases. The general relationship of the Aryan languages to their common ancestor is, however, no less clearly manifest than that of the modern Romanic languages to the Latin. After fifty years of such comparative study, in a cautious and prudent way, we have succeeded in making out some few cases of demonstrable genetic kinship among groups of languages. Beside the Aryan family, in the study of which such profound knowledge has been obtained, we have clearly made out the existence of the Dravidian family in Southern India, and of the Altaic family, — to which the Finnish, Hungarian, and Turkish belong, — to say nothing of the long-established Semiticfamily. Other families of speech no doubt exist, and will by and by have their relationships definitely marked out. But the moment we try to compare these families with each other, in order to detect some definable link of relationship between them, we are instantly baffled. Any true family of languages will show a community of structure as conspicuous as that which is seen among vertebrate animals. The next family you study will be as distinctly marked in its characteristics as is the group of articulated insects, spiders, and crustaceans But to compare the two families with each other will prove as futile as to compare a reindeer with a lobster. The only conclusion to which you can logically come is that while certain languages, here and there, have become variously modified, so as to give rise to well-defined families of speech, the like process has not taken place universally. In other words, the derivation of a dozen languages from a common ancestor is not a permanent and universal, but a temporary and local phenomenon in the history of human speech, and we need not expect to come across any such fact of derivation, except where it can be duly accounted for by the peculiar circumstances of the case.
This conclusion is reinforced when we consider the circumstances under which a single language gives rise to several mutually resembling descendants. Obviously such a language must have a high degree of permanence and a wide extension. It must be spoken for a long time by large bodies of men spread over a wide territorial area. Take, for example, the rise of the modern Romanic languages from the Latin. In the fourth century after Christ the Latin language was spoken all over the Italian and Spanish peninsulas, throughout most of Gaul and Switzerland, along the banks of the Upper Danube, and in what are now called the Roumanian principalities. In all these countries Latin was the speech in which the ordinary affairs of life were transacted, and this had come to be so mainly because the native dialects of these countries were numerous and uncultivated ; and as all were in close political and social connection with Rome, it was a much simpler matter for all to learn Latin than for the Romans and their subjects alike to learn a score of barbarous tongues. The business of life got more easily transacted in this way. No such result followed the conquest of the Eastern world, because Greek was spoken all over the East, and every educated Roman knew Greek already ; so that in this case it was a simpler matter for the conquerors to talk Greek than for their subjects to learn Latin, Practical convenience is the final arbiter in pretty much all such cases. Now it must not be supposed that the Latin talked all over the West was quite like the elegant language of Cæsar and Virgil. It was only educated people in Rome or Milan, and perhaps in such cities as Nismes or Lyons, that talked like this. Colloquial Latin always had plenty of dialectic peculiarities. Even in Italy the Latin had supplanted, in former times, a number of kindred Umbrian and Sabine dialects, and we may be sure that all these left their mark upon the common speech. In getting diffused over Europe, this impure colloquial Latin could not fail to pick up here and there some peculiar word or phrase, while now and then some other word or phrase would be lost from its old stock and forgotten, so that people did not talk just alike throughout the empire. A Spaniard’s local peculiarities of utterance and phraseology were distinguishable from those of a Rhætian. though both talked Latin and could understand each other.
Now as every language changes more or less from age to age, so the speech of the Romans in the fourth century after Christ had come to differ in many respects from the speech of their forefathers who, six hundred years earlier, had fought against Hannibal. But up to this time the intercourse between the various parts of the Roman world had been so close and continuous that the capital still furnished the standard of discourse for the whole empire. During the next six centuries a different set of circumstances was at work. For a second time the Latin language was learned by scores of barbarous tribes, but this time it was no longer Rome that set the fashion and maintained the standard. In innumerable provincial towns and barbaric assemblies new standards of speaking were gradually established. The lines of connection, administrative and commercial, which had formerly been kept up, were in many cases severed, and each little tract of country led a more sequestered life than before. Many new expressions came into use, — Teutonic in Gaul and Italy, Arabic in Spain, Slavic in Roumania; and local idioms and peculiarities of accent multiplied, in the absence of a uniform standard. In this way the vulgar Latin insensibly diverged into a host of provincial dialects, or patois, the divergence being great or little according to the frequency of intercourse between different localities. Thus the Tuscan and the Savoyard could both understand the Milanese, the inhabitant of Lyons could talk with the Savoyard and with the citizen of Orleans, and the Orleanese would be intelligible to the Parisian ; while, on the other hand, the Parisian could hardly carry on a conversation with the Savoyard, and would be quite incapable of understanding the Tuscan. Some such slowly-graded transition may still be noticed by the traveler from France to Italy who takes pains to observe the speech of the common people. At Nice, for instance, local newspapers are published in a dialect which one hardly knows whether to call French, Provençal, or Italian.
After this process of divergence had gone on for some time, a new start was taken toward uniformity, but in such a way as to enhance and complete the divergence already begun. When literary men gave up trying to write classical Latin, and began to clothe their thoughts in the colloquial Romance or vulgar tongue of the times, new centres of political and intellectual life had begun to be formed at Paris, Toulouse, and Florence ; and the dialects of these cities began to assume preëminence as literary and fashionable dialects. As Southern France came more and more under the sway of Paris, the second of these centres indeed lost its relative importance, and the Provençcal tongue
gradually sank into an unfashionable patois; but Parisian and Tuscan, on the other hand, came to be so generally read and spoken that after a while they quite crowded their intermediate sister dialects out of sight, and to-day they are the sole recognized representatives of good French aud good Italian speech, although there is still a great deal of French spoken that is not Parisian, and a great deal of Italian that is not Tuscan. This predominance of the two central dialects is in our day increasing more rapidly and decisively than ever before, and the process will unquestionably go on until all Frenchmen speak Parisian, and all Italians speak Tuscan. Railroads and telegraphs, newspapers and novels, have already sealed the death-warrant of all patois, and the execution is only a question of time. It is because of the wide diffusion in our own country of these powerful agencies for keeping men in contact with each other that we have no varieties of dialect here worth speaking of. It is not at all likely that in this country such dialectic variations will ever spring up. And for the same reason it is not likely that any essential divergence will ever arise between the English language as spoken in England and the same language as spoken in America. In the Middle Ages, wolves, brute and human, taxgatherers, and robber barons, as well as bad roads and imperfect vehicles, made a few miles of wood or mountain a greater barrier to intercourse than the wide ocean is to-day. For the language of the thriving people to whom, as to the ancient Greeks, the ocean has become (πóυτος) a common “ pathway ; ” who have taught mankind how to drive ships with steam, and how to send electric flashes of intelligence through the watery abyss, — for this language a future of unprecedented glory is in store. By the end of the twentieth century, English will no doubt be spoken by something like eight hundred million people. crowding all over North America and Australia, as well as over a good part of Africa and India, with island colonies in every sea, and naval stations on every cape. By that time so large a proportion of the business of the world will be transacted by people of English descent that, as a mere matter of convenience, the whole world will have to learn English. Whatever other language any one may have learned in childhood, he will find it necessary to speak English also. In this way our language will become more and more cosmopolitan, while all others become more and more provincial, until, after a great length of time, they will probably one after another assume the character and incur the fate of local patois. One by one they will become extinct, leaving English as the universal language of mankind.
There is, I think, a considerable probability that things will come to pass in this way, though the process must of course be a very slow one, and the result here prefigured will very likely come so far down in the future as to coincide with the disappearance of barbarism from the earth, and with the inauguration of that pacific “ parliament of man ” of which the philosophic poet has told us. But, however the actual result may shape itself in its details, the considerations here brought forward would seem to indicate that complete community of speech belongs rather to the later than to the earlier stages of human progress. What we may regard as certain is that community of speech on a wide scale requires prolonged and continuous business communication among large bodies of men. Where communication is seriously interrupted for a long period of time, as in the Dark Ages of Europe, the tendency is for the common language to break up into a number of more or less similar dialects; and in proportion as frequent communication is resumed there is manifested an opposite tendency of a few central dialects to crush out their neighbors, and to grow into wide-spread languages. This is, in brief, the way in which languages grow, and diverge, and supplant one another. There is nothing that is mysterious or metaphysical in the process ; it is purely a matter of practical convenience. In the long run the actions of man are determined by what we may call the “ law of least effort: ” the easiest way of doing things is the one which, sooner or later, is sure to be adopted ; and to this general law the myriad little actions involved in speech form no exception.
Carrying back to ancient times the lesson we have learned from the career of Latin, we find that the facts, so far as known, sustain our conclusion. Among the Semitic peoples there was undoubtedly a time when all were of one blood and one speech. No one doubts that Arabs, Jews, and Syrians are as closely related by descent as Germans, Swedes, and Englishmen. The social condition of these Semitic races, shortly before the historic period, is best represented by the wandering Arabs of the present day. In this patriarchal stage of society there is no such close political cohesion as there is among nations of modern type, but there is frequent intercourse for business purposes, and even sometimes for purely literary objects, as in the old competitions of bards at Mecca before the time of Mohammed ; and this intercourse has sufficed to preserve the main features of the language. In early times there was sufficient communication between the patriarchal tribes of Arabia and Palestine and the adjacent civilized nations of Assyria, Babylonia, and Phœnicia to prevent any very wide divergence of speech. The differences between Hebrew, Syriac, and Assyrian are not greater than the differences between French, Spanish, and Italian.
So, too, in the direct line of our own ancestry, we find that the primitive Aryans were a race partly agricultural and partly pastoral in pursuits, living in durable houses, grouped together into large villages, surrounded by defensible walls. The structure of the family was somewhat cruder than among the patriarchal Arabs and Hebrews ; the social and political system was such as we see vestiges of to-day in the village communities of Russia and Hindustan. Preeminent among all early races in the rearing of flocks and herds, the old Aryans required immense grazing grounds, and would seem to have occupied all the wide grassy plains which lie between the mountains of Central Tartary and the southern slopes of European Russia. At the same time their agricultural pursuits and their durable villages imply a considerable amount of political stability, and there is good evidence that for a long time a common language was spoken throughout this vast territory. As we follow these Aryan tribes in their great career of permanent conquest and settlement, one branch into Persia and India, and other branches into Greece, Italy, Germany, Gaul, and Britain, we come upon the same linguistic phenomena which we observed above in the mediæval history of Latin. With the isolation of the various tribes, separated from each other by wide distances, we see the Aryan mother-tongue break up into innumerable dialectic forms ; until, by and by, with the rise of new and distinct centres of social life, new and distinct languages come upon the scene, and acquire literary immortality in the Vedas, in the Avesta, in the epics of Homer and Virgil, in the novels of Cervantes and Turgenief, in the sermons of Bossuet and Taylor, in the dramas of Shakespeare and Goethe, and in that palladium of linguistic stability in the future, — the English version of the Bible.
In such cases as these, where a single durable mother-language has produced several durable offspring, the signs of kinship, whether in grammar or in vocabulary, are never obliterated. After an independent career of more than ten centuries, the genetic relationship of French and Italian is a perfectly patent fact, about which there could be no question whatever, even if all memory of the Roman Empire had lapsed from men’s minds, even if some fanatical Cardinal Ximenes had burned in a bonfire every scrap of French and Italian literature that ever existed. After an independent career of not less than forty centuries, the kinship of Latin and Sanskrit is equally unmistakable. It is not an occult fact, which discloses itself only after a subtle philological analysis ; it is a fact so plain that no one who reads Sanskrit and Latin books can possibly overlook it, and it forced itself upon the attention of the first European scholars who studied Sanskrit in the seventeenth century, though they knew nothing of philological analysis as we understand it. The similarity between the longknown Hebrew and the lately-deciphered Assyrian is no less conspicuous; and the same may be said of the Dravidian languages of Southern India when compared with one another.
But as we leave this circle of studies, and venture out into the wilderness of barbaric speech, we find a very different state of things. The northern portions of Asia have been inhabited, within the period of history, by three different races, all of whom still survive, — the Finno-Tataric, the Mongolian, and the Samoyedic races. The linguistic relationships of these peoples are very instructive. In the first place, the Finno-Tataric peoples appear to belong to the same white race from which the Aryans and the Semites have diverged, although there is nothing remotely resembling Aryan or Semitic in FinnoTataric speech. This family of languages is represented in Europe by the Finnish and its neighboring dialects, by the Hungarian, and by the Turkish. In Asia it is represented by a great number of languages, spoken in the Caucasus, in Turkistan, and in Siberia. Eastward of this vast region comes the Mongolian or yellow race, with which we should be very careful not to confound the Tatars. There has always been a great deal of confusion of nomenclature in speaking of these races, but the lines of distinction are really simple enough when we have once learned them. The ambiguous word which is responsible for most of the confusion is the epithet Tatar, which did originally belong to the Mongols, but has come to be applied by preference to the Turkish family. When Jinghis Khan, in the thirteenth century, made the name Tatar a sign of terror and humiliation to all Asia and Europe, it became customary to apply this dreaded epithet to all the hordes that were subject to the Mongolian ruler, — changing the word slightly to “ Tartar,” so as to add to it a mild flavor of the bottomless pit, in allusion to the general behavior of those ugly customers. As most of these hordes with which Europeans came into contact were really of white or Turkish race, the name Tatar became gradually appropriated to these, and thus became unfit for distinguishing the yellow Mongolians. All ambiguity would be avoided if we were to drop the name Tatar altogether, and substitute the name Turk for the whole group of peoples of which the Ottomans are the most conspicuous. Our school atlases already have “ Turkistan ” instead of the old-fashioned “ Independent Tartary.”
The Mongolian race comprises the yellow tribes of Central Asia, from whom came Jinghis Khan, Timur, and the whole line of Mogul sovereigns of India; and also the Tungusians, or Mandshus, who for the last two centuries have ruled over China. The Chinese themselves, as well as the Japanese, must also be considered as branches of the Mongolian race. On the other hand, the Samoyeds of Northern Siberia seem to be allied to our Eskimos, but not very obviously to the Mongolians.
The race divisions of the northern half of Asia are thus clear enough. First, we have the Finno-Tatars, or FinnoTurks, belonging to the dark-haired portion of the great white race ; secondly, we have the Mongolians ; thirdly, the arctic Samoyeds. But the languages spoken by these peoples cannot be classified in any such simple way. The languages of the Finns and Turks carry us back to two mother-tongues, and these are possibly reducible to one. It is otherwise when we come to Mongolian speech. On the one hand, the Mongolian dialects of Central Asia are strikingly similar in structure to the Tungusian languages, and also to the Japanese; and in these structural peculiarities they agree also with the Finno-Turkic. On the other hand, when we study the vocabularies, we do not find any similarity, such as to suggest a primitive identity, between Japanese, Tungusian, and Mongolian proper. We are still further baffled when we come to Chinese. The people of Japan obtained their written character from China, modifying it to suit the needs of their own language ; and so a Japanese printed page looks very like a printed page in Chinese. If you were just to look at these printed pages, you would imagine that the two languages are very similar, just as a Chinaman, on seeing Hungarian printed in the Roman character, would fancy that Hungarian must be similar to English or Latin. In reality no kinship has yet been detected between the languages of China and Japan. Not only in vocabulary does Chinese differ from all the other languages spoken by the Mongolian race, but it even presents a fundamentally distinct type of linguistic structure. Age after age, from the remotest antiquity to which historic or philologic inference can guide us, the Chinese have talked with different words and after a different grammatical fashion from their yellow neighbors ; and these in turn have maintained each their distinct varieties of speech; although all these peoples — the inhabitants of Japan and China, the Tungusians, and the Mongols of Central Asia — are undoubtedly united by physical bonds of descent from one and the same primeval yellow race.
The inference from this is that there never was a primitive Mongolian mother-tongue in the sense in which there was a primitive Aryan mother-tongue. The common ancestors of Japanese, Chinese. Tungusian, and Mongol never at any time lived together in one great society, welded into a unit by community of language, traditions, and customs, as was the case with the common ancestors of Roman, Teuton, and Hindu. On the contrary, the aboriginal yellow men must have roamed about in detached tribes, like the blacks of Australia or the red men of America, with half-formed languages fluctuating from generation to generation, diverging with great rapidity, and speedily losing all traces of their origin. Ensconced within convenient mountain barriers, one series of these yellow tribes worked out its peculiar language and civilization in the rich hillcountry and along the great navigable rivers of China. A second series of tribes, moving without reference to these, and probably at a much later date, formed a permanent community in the islands of Japan. While the remainder of the race have led a nomadic life down to the present day; now and then engaging in combined activity for a generation or two, under the guidance of such adventurers as Attila, or Jinghis, or Timur, to become for a brief season the “ scourge of God" and the terror of mankind, but ever, as now, incapable of stable political union. With such divergent careers as these, we need not expect to find evidence of linguistic community among the different branches of the yellow race. If we find one set of linguistic phenomena in China, and a totally different set in Japan, and yet another set among the barbarous Mongols and Tunguses, this is no more than we might have expected. We need not expect to find such phenomena as the coördinate divergence of French and Italian from a common Latin mother-tongue, or of Latin and Sanskrit from a common Aryan mother-tongue, except where we can find historical conditions similar to those under which these phenomena were manifested. Outside of that broad stream of history which includes the Aryan and Semitic worlds we do not find such conditions, save in a few sporadic cases. On the contrary, we find just such a state of things as would follow from the isolated independent development of a number of languages, either without any original kinship, or with the original kinship blurred and destroyed almost from the very beginning.
The last clause introduces us to a consideration concerning barbarous languages which is of the first importance. There is a certain sense in which we may admit community of origin for languages that are now quite dissimilar; but the sense is one that is foreign to philological usage, and has no real philological significance. No doubt all the yellow races of Asia are descended from some small group of yellow progenitors, and no doubt this ancestral group possessed the faculty of articulate speech. Most likely the group was at the outset small enough to use but one language, and as the group increased in size and became subdivided into a number of tribes, the common language would soon get broken up into dialects. So far very good; but what we have to notice is that under such circumstances the breaking up of the common language would not in any way resemble the breaking up of Latin into the dialects of France and Italy. On the contrary, the several dialects would change so rapidly as to lose their identity : within a couple of centuries it would be impossible to detect any resemblance to the language of the primitive tribe. The speech of uncivilized tribes, when not subject to the powerful conservative force of widespread custom or permanent literary tradition, changes with astonishing rapidity. Such languages usually contain but a few hundred words, and these are often forgotten by the dozen and replaced by new ones even in the course of a single generation. Among many South American Indians, as Azara tells us, the language changes from clan to clan, and almost from hut to hut, so that members of different families are obliged to have recourse to gestures to eke out the scanty pittance of oral discourse that is mutually intelligible. In the northern part of Celebes, “in a district about one hundred miles long by thirty miles wide, not less than ten distinct languages are spoken.” 1 In civilized speech no words stick like the simple numerals : we use the same words to-day, in counting from one to ten, that our ancestors used in Central Asia ages before the winged bulls of Nineveh were sculptured ; and the change in pronunciation has been barely sufficient to disguise the identity. But in the language of Tahiti five of the ten simple numerals used in Captain Cook’s time have already become extinct: —
“ Two was rua ; it is now piti. Four was ha; it is now meha. Five was rima ; it is now pae. Six was ono ; it is now fene.
Eight was vara ; it is now vau.”2
Out of many facts that might be cited, these must suffice. The facility with which savage tongues abandon old expressions for new has no parallel in civilized languages, unless it be in some of the more ephemeral kinds of slang. It is sufficiently clear, I think, that under such circumstances a language will seldom or never acquire sufficient stability to give rise to mutually resembling derivative dialects. If the habits of primitive men were in general similar to those of modern savages, we need not be surprised that philologists are unable to trace all existing languages back to a common origin. In order to get back to a universal mother-tongue, it would almost seem requisite that the history of mankind should have begun with universal empire.
We shall conclude, I think, after a survey of the whole matter, that in speech, as in other aspects of social life, the progress of mankind is from fragmentariness to solidarity : at the beginning, a multitude of feeble, mutually hostile tribes, incapable of much com bined action, with hundreds of halfformed dialects, each intelligible to a few score of people; at the end, an organized system of mighty nations, pacific in disposition, with unlimited reciprocity of intercourse, with very few languages, rich and precise in structure and vocabulary, and understood by an men.