Language and the Study of Language/Lecture VI

by William Dwight Whitney

LECTURE VI.

Languages and literatures of the Germanic, Slavonic, Lithuanic, Celtic, Italic, Greek, Iranian, and Indian branches of Indo-European speech. Interest of the family and its study; historical importance of the Indo-European races; their languages the basis of linguistic science. Method of linguistic research. Comparative philology. Errors of linguistic method or its application.

Our consideration of the processes of linguistic growth, and of their effects upon the condition of language and the rise of discordant tongues, was brought to a close in the preceding lecture with a brief discussion of certain erroneous views respecting original dialectic variety, and the influence exerted upon it by literary and grammatical cultivation. We then looked to see how and how far the principles which we had established could be applied to explain the seemingly infinite confusion of tongues now prevailing upon the earth, and to facilitate their classification and reduction to order. This led us to a recognition of our own language as one of a group of nearly related dialects, the Germanic group; and, on inquiring farther, we found that this was itself a member of a wider family, embracing nearly all the tongues of Europe, with a part of those of Asia, and . divided into seven principal branches: namely, the Indian, the Iranian, the Greek, the Latin, the Germanic, the Slavonic (including the Lithuanic, sometimes reckoned as a separate branch), and the Celtic. We called it the Indo-European family. At some place and time, which we were obliged to confess ourselves unable to determine with any even tolerable degree of confidence—but more probably in Asia, and certainly not less than three thousand years before Christ—and in a condition of civilization respecting which the evidence of language furnished us valuable hints, some single community had spoken a single tongue, from which all these others were descended, in accordance with the universal laws of linguistic tradition, by processes which are still active in every part of human speech. And now, waiving for a while the question whether it may not be possible to regard the great Indo-European family itself as only a member of a yet vaster family, including all or nearly all the languages of the human race, we have, in the present lecture, to review more in detail its constitution, to note the period and locality of its constituent members, to glance at the special historical importance attaching to them and to the peoples who speak them, to set forth their value as the fundamental material of linguistic science, and to examine anew and more systematically the general method of linguistic research, as established upon their study.

We may best commence our survey of the varieties of Indo-European speech with our own branch, the Germanic. Its existing dialects, as has been already pointed out, are divided into three groups or sub-branches: 1, the Low-German, occupying northern Germany and the Netherlands, with their colony Britain, and with the numerous and widely-scattered modern colonies of Britain; 2, the High-German, in central and southern Germany; 3, the Scandinavian, in Denmark, Sweden, Norway, and Iceland. Of the Low-German group, the English is by far the most important member; its eventful history, illustrated at every step by valuable literary documents, we trace back, through Middle English (A.D. 1350-1550), Old English (A.D. 1250-1350), and Semi-Saxon (A.D. 1150-1250), to the Anglo-Saxon, which reaches into the seventh century of our era, possessing an antiquity exceeded by only one other Germanic dialect. Its earliest monuments, in their style and metre, and at least one of them, the Beowulf, in subject and substance also, carry us back to the pre-Christian period of Germanic history. We cannot delay here to enter into any detailed examination of the character and changes of English speech, interesting and instructive as such a task would be; save so far as they have been and may hereafter be brought in by way of illustration of general linguistic laws, they must be left to more special treatises.^[1]

Next of kin with the Anglo-Saxon, or oldest form of English, are the ancient Frisian, of the northern sea-coast of Germany, which had, in the fourteenth century and later, a literature of its own, of juridical content, composed in an idiom of form little less antique than Old High-German, notwithstanding its comparatively modern date—and the Old Saxon, the principal language of northern Germany between the Rhine and the Elbe, represented to us by but a single work, the Heliand or 'Savior,' a poetical life of Christ, probably of the ninth century. Both Saxon and Frisian have been almost wholly crowded out of cultivated use in modern times, as was explained in a former lecture (see p. 164), by the overpowering influence of the High German, and their domain has also been encroached upon by other dialects of the same kindred, so that they survive at present only as insignificant popular patois. Nothing but the political independence of Holland has saved its peculiar speech from the same fate: the literary cultivation of the Netherlandish or Dutch can be traced back to the thirteenth century, although dating chiefly from the sixteenth, the era of the country's terrible struggle against the political tyranny of Spain. The Flemish, the closely allied idiom of Flanders, has its own separate records, of about the same antiquity, but is now nearly extinct.

The history of High-German speech was succinctly sketched in connection with our inquiries into the rise and extension of literary dialects. It falls into three periods. The first period is that of the Old High-German (Althochdeutsch), from the eighth to the twelfth century; its monuments are tolerably abundant, but, with trifling exceptions, of Christian origin and religious content: they represent three principal sub-dialects, the Frankish, the Alemannic and Swabian, and the Bavarian and Austrian. The second period, that of the Middle High-German (Mittelhochdeutsch), covers about four centuries, beginning with the twelfth and ending with the fifteenth; its ruling dialect is the Swabian; and its rich literature hands down to us valuable productions of the poetical fancy of the times, in the lyric verses of the Minnesingers, and precious memorials of ancient German national tradition, in the heroic legends (Heldensagen). The foremost work of the latter class, the Lay of the Nibelungen (Nibelungenlied), is one of the noblest epics which any country has produced, in any age of the world. Of the language and literature of the New High-German period, from early in the sixteenth century to our own times—the "German" language and literature, as we are accustomed to call it—there is no need that I speak more particularly.

The third subdivision of the Germanic branch is the Scandinavian. Its earliest monuments come to us from Iceland, that far-off and inhospitable island of volcanoes, boiling springs, and ice-fields, which, settled in the ninth century by refugees from Norway, long continued a free colony, a home of literary culture and legendary song. Cbristianity, more tolerant there than elsewhere on Germanic soil, did not sweep from existence the records of ancient religion and customs. The two Eddas, gathered or preserved to us from the twelfth and thirteenth centuries, are, in virtue of their tone and content, by far the most primitive works in the whole circle of the Germanic literatures, documents of priceless value for the antiquity of the Germanic race. Their language also, though of so much more recent date than the oldest Anglo-Saxon and High-German, is not exceeded by either in respect to the primitiveness of its phonetic and grammatical form. Nor has it greatly changed during the six or seven centuries which have elapsed since the compilation of the Eddas. The modern Icelandic is still, among all the existing Germanic tongues, the one that has preserved and possesses the most of that original structure which once belonged to them all alike. Three other dialects, the Norwegian, the Swedish, and the Danish, constitute along with it the Scandinavian group, and are languages of literary culture. They are not direct descendants of the "Old Norse" tongue, as the ancient Icelandic is usually called: the Norwegian comes nearest to being so; the others represent more ancient dialectic divisions of Scandinavian speech.

How many other Germanic branches, originally coördinate with the three we have described, once had existence, but have become extinct in later times, by the extinction of the communities who spoke them, we have not, nor shall we ever have, any means of knowing. But of one such, at least, most precious remains have escaped the general destruction of the nationality to which it belonged. One portion of the western division of the great and famous Gothic nation crossed the lower Danube, some time in the early part of the fourth century, and settled in the Roman province of Mœsia, as subjects of the empire and as Christians. For them, their bishop and leader, Ulfilas, later in the same century, made a translation into their own vernacular of nearly the whole Bible, writing it in an alphabet of his own devising, founded on the Greek. Five hundred years afterward, the Gothic was everywhere an extinct tongue; but considerable portions of the Gothic Scriptures—namely, a part of the Gospels, Paul's epistles nearly complete, and fragments of the Old Testament—are happily still preserved, in a single manuscript of the fifth century, now at Upsala, in Sweden. Scanty as these relics may be, they are of inestimable value in illustrating the history of the whole Germanic branch of Indo-European language, and bridging over the distance which separates it from the other branches. For, as in time, so still more notably in material and structure, their idiom is much the most ancient of all the varied forms of Germanic speech: it is not, indeed, the mother of the rest, nor of any among them; but it is their eldest sister, and fully entitled to claim the place of head of their family.

The Slavonian branch—to which, on account of its local vicinity, as well as its probable nearer relationship, to the Germanic, we next turn our attention—need not occupy us long. It is of much less interest to us, because of its greater remoteness from our race and from our knowledge, its inferior historical importance and literary value, and its more modern appearance.^[2] The oldest of its dialects in date, and, in nearly all respects, the most primitive in form, is the language of the ancient Bulgarians, into which their apostle Cyril translated the Scriptures, now just about a thousand years ago. It is a curious coincidence that our knowledge of both Germanic and Slavonic speech thus begins, like that of many a rude and hitherto unlettered dialect in the hands of missionaries at the present day, with a Bible version, and at nearly the same geographical locality; the kingdom of the Bulgarians having followed that of the Goths on the southern bank of the lower Danube. But this ancient idiom—from which the modern Bulgarian differs greatly, having changed with unusual rapidity in the interval—is more commonly called the Old Slavonic, or the Church Slavic, having been adopted by a large part of the Slavonian races as their sacred language, and being still employed as such, within the ecclesiastical limits of the Greek Church. It belongs to what is known as the south-eastern section of the Slavonic branch. By far the most important of the other languages in the same section is the Russian, in its two divisions, the Russian proper and the Little-Russian, or Ruthenian. The Russian is in our day a literary language of considerable importance; its forms are traceable, in scanty documents, back into the eleventh century. In its cultivated development, it has been strongly influenced by the Church Slavonic. The south-eastern section further includes the Servian, with its closely related dialect, the Kroatian, and the Slovenian of Carinthia and Styria. Specimens of these tongues are as old as the tenth, or even the ninth, century. The Servian has an interesting modern literature of popular songs.

To the other section, the western, belong the Polish, the Bohemian with the related Moravian and Slovakian, the upper and lower Sorbian, and the Polabian, on the Elbe. Of these, the Bohemiam is the oldest, having monuments probably of the tenth century. Polish literature begins in the fourteenth century, since, down to that time, the cultivated of the nation had written wholly in Latin. The others can show nothing older than the sixteenth century, and are of little consequence in any aspect.

The Lithuanic or Lettic group of dialects is sometimes treated as a subdivision of the Slavonic, and sometimes—perhaps with better reason—as a separate branch, coördinate with the other, although very closely related to it. It is of very slight historical or literary importance: its interest lies chiefly in the fact that, under the operation of causes in its history which are yet unexplained and probably unexplainable, it has preserved many of the original forms of Indo-European speech in a more uncorrupted condition than any other known dialect of the whole family which is not as much as two thousand years older. It is composed of only three dialects, one of which, the Old Prussian, the original language of the inhabitants of north-eastern Prussia, has been extinct for two hundred years, crowded out of existence by the Low-German, and leaving behind, as its only monument, a brief catechism. The other two, the Lithuanian and the Lettish, or Livonian, are still spoken by a million or two of people in the Russian and Prussian provinces bordering on the Baltic, but seem destined to give way helplessly before the encroachments of the German and Russian, and to share one day the fate of their sister-dialect. The oldest Lithuanian document dates from the middle of the sixteenth century. The southern or High Lithuanian is of most antique form; the Low Lithuanian, and yet more notably the Lettish to the north, show a less remarkable conservation of ancient material.

The Celtic languages, as was pointed out in the last lecture, have been well-nigh extinguished by the Romanic and Germanic tongues, and now only lurk in the remotest and most inaccessible corners of the wide territory which they once occupied in Europe. The Scotch Highlands, the wildest parts of Ireland, the Isle of Man, the mountains of Wales, the rough glens of Cornwall, and the land lying nearest to Cornwall across the British Channel, the promontory of Brittany, are the only regions where, for many centuries past, Celtic speech has been heard. The Cornish, too, has become extinct within the memory of the present generation; the Irish is rapidly on its way to the same fate; the Gaelic will not survive the complete taming and civilization of the Highlands; the French is likely to crowd out the patois of the Breton peasant; and it is greatly to be doubted whether even the Welsh people, passionate as is the attachment with which at present they cling to their peculiar speech, will continue always to refuse the advantages that would accrue to them from its relinquishment, and a more thorough fusion with the greater community of speakers of English to which they form an adjunct. There has never been a homogeneous, independent, and cultivated Celtic state, capable of protecting its idiom from the encroachment of other tongues; and only such protection, now unattainable, can, as it seems, save Celtic speech from utter extinction.

There is no small difficulty in treating satisfactorily the documents which illustrate the history of the Celtic languages, owing to the prevalence of a peculiar and strongly-marked linguistic disease, well known among philologists as "Celtomania," which has been very apt to attack students of the subject—especially such as were of Celtic extraction, but in some degree foreigners also—leading them wildly to exaggerate the antiquity and importance of the Celtic civilization, language, and literature. We have had Celtic set up as the most primitive and uncorrupted of tongues, spoken by generations long anterior to the oldest worthies whom history, sacred or profane, recognizes, and furnishing the only sure foundation to universal etymology; we have had ancient inscriptions and difficult texts, of the most diverse origin and distant locality, explained out of Celtic into high-sounding phrases, of true Ossianic ring; we have had the obscure words of various languages traced to Celtic roots, provided with genealogies from an Irish or Welsh ancestor—and much more of the same sort. Sober and unprejudiced inquiry cuts down these claims to greatly reduced, though still respectable, dimensions.

So completely were the Gaulish dialects of northern Italy, France, and Spain wiped out by the Latin, so few traces of them are left to us, either in the later idioms of the Latin or in fragments of writings, inscriptions, and coins, that it is still a matter of doubt and question among Celtic scholars to which of the known divisions of Celtic speech, the Gadhelic or the Cymric, they belonged, or whether they did not constitute a third division, coördinate with these. Aside from the exceedingly scanty and obscure Gallic epigraphical monuments, and the few single words preserved in classic authors, the earliest records, both of Irish and Welsh speech, are glosses, or interlinear and marginal versions and comments, written by Celtic scholars upon manuscripts which they were studying, in old times when Wales and Ireland, especially the latter, were centres of a lively literary and Christian activity. Of these glosses, the Irish are by far the most abundant, and afford a tolerably distinct idea of what the language was at about the end of the eighth century. There is also an independent literary work, a life of Saint Patrick, which is supposed to belong to the beginning of the ninth century. The other principal Gadhelic dialect, the Scotch Gaelic, presents us a few songs that claim to be of the sixteenth century. The Ossianic poems, which excited such attention a hundred years ago, and whose genuineness and value have been the subject of so lively discussion, are probably built upon only a narrow foundation of real Gaelic tradition.

In the Cymric division, the Welsh glosses, just referred to, are the oldest monuments of definite date. Though hardly, if at all, less ancient than the Irish, coming down from somewhere between the eighth and the tenth centuries, they are very much more scanty in amount, hardly sufficient to do more than disprove the supposed antiquity of the earliest monuments of the language that possess a proper literary character. For long centuries past, the Welsh bards have sung in spirit-stirring strains the glories and the woes of their race; and it is claimed that during much more than a thousand years, or ever since the sixth century, the era of Saxon invasion and conquest, some of their songs have been handed down from generation to generation, by a careful and uninterrupted tradition. And the claim is probably well founded: only it is also pretty certain that, as they have been handed down, they have been modernized in diction, so that, in their present form, they represent to us the Welsh language of a time not much preceding the date of the oldest manuscripts, or of the twelfth to the fourteenth centuries. The later Welsh literature, as well as the Irish, is abundant in quantity. The Cornish, also, has a tolerably copious literature of not far from the same age; its earliest monument, a Latin-Cornish vocabulary, may be as old as the twelfth century. The language of Brittany, the Armorican—which is so closely allied with the two last-mentioned that it cannot well be regarded as a remnant and representative of the Celtic dialects of Gaul, but must rather belong to colonists or fugitives from Britain—is recorded in one or two brief works going back to the fourteenth century, or even farther.

We come next to the Romanic branch, as we have called it when briefly noticing its history at an earlier point in our discussions. Of the languages which compose it, and whose separate currents of linguistic tradition we trace backward until they converge and meet in the Latin, two, the Rhæto-Romanic in southern Switzerland and at the head of the Adriatic, and the Wallachian of the northern provinces of Turkey, have no literature of any antiquity or independent value. The other five—the Italian, French, Provençal, Spanish, and Portuguese—all emerged out of the condition of vulgar patois, and began to take on the character of national cultivated languages, at not far from the same time, or in the eleventh, twelfth, and thirteenth centuries. There are fragments of French texts dating from the tenth century, but the early French literature, abundant and various, and, in its romances, attaining a wonderfully sudden and general popularity throughout cultivated Europe, belongs to the twelfth and thirteenth centuries. The Provençal poetry, consisting of the songs of the troubadours, whose chief activity was displayed at the court of Toulouse, in southern-most France, was wholly lyrical in form, and amatory or satirical in content: it finished its brilliant but brief career, of about three hundred years, in the fourteenth century. The culture of Italian begins at the court of Frederic II., about A.D. 1200, and within a century and a half of that time lived, sang, and narrated the three greatest writers of Italy—Dante (ob. 1321), Petrarch (ob. 1374), and Boccaccio (ob. 1375). The Spanish heroic legend commences in the twelfth century; and there are monuments of Portuguese speech of about the same time. Among these languages, the French is that which has undergone most change during the historical period; the oldest French and Provençal form a kind of middle term between the modern language and the ancient Latin, illustrating the transition from the latter to the former.

But if we have called the branch of Indo-European speech to which these tongues belong the Romanic, we have done so out of regard to its later history and present constitution, and not altogether properly. To the student of Indo-European philology, these are the recent branchings of a single known stock, the Latin; to trace their development is a task of the highest interest, a whole linguistic school in itself; they furnish rich and abundant illustration of all the processes of linguistic growth: but, as regards any direct bearing upon the history of Indo-European speech, they have value only through the Latin, their common parent. The remoter relations of the Latin itself receive light from various sources. In its familiar classic form, it represents to us the speech of the learned and educated Romans of a century or two before the Christian era; it is somewhat refined by literary culture from the diction of the oldest authors whose works have come down to us, in fragments or entire—as Livius Andronicus, Plautus, Terence—and is far more notably changed from the language of earlier Roman times—as is shown by the yet extant monuments, like the inscription on the Duilian column (about B.C. 260), that on the sarcophagus of a founder of the Scipio family (a little older than the last mentioned), and especially the Salian hymn and song of the fratres arvales, of yet earlier but uncertain date, in which the best Latin scholar would find himself wholly at fault without the traditional interpretation which is handed down along with them: in these monuments is preserved to us many an antique form, giving valuable hints respecting the grammatical and phonetic development of the language. Their evidence is supplemented in a very important manner by that of other kindred Italian dialects. The Oscan or Opican of southern Italy was the language of the Samnites and their allies, from whose hands Rome wrung after a severe and often doubtful struggle the dominion of the peninsula: it was not disused as the official speech of some of the southern provinces until less than a hundred years before Christ; and coins and inscriptions dating from the two or three preceding centuries still teach us something of its structure and character. The Umbrian, the tongue of north-eastern Italy, is yet more fully represented to us by the Euguvine tablets, inscribed with the prayers and ceremonial rules of a fraternity of priests, and supposed to be as old as the third and fourth centuries before our era. Of the Volscian dialect, also, and the Sabine or Sabellian—the former being more akin with the Umbrian, the latter with the Latin—some exceedingly scanty relics have been discovered. The interpretation and comprehension of all these—resting, as it does, solely upon comparison with the Latin and other more distantly related tongues—is at present, and is likely always to remain, incomplete and doubtful; but they are of essential importance, both in explaining some of the peculiarities of the Latin, and in fixing its position as one of a group of kindred dialects occupying the greater portion of the Italian peninsula, and hence most suitably to be denominated the Italic group. The theory that the Latin was produced by a mixture of somewhat discordant elements—of Roman, Sabine, and Oscan; or of these and Etruscan—brought together by historical circumstances, and finally fused into homogeneousness, is one which belonged to a former stage of linguistic science, and is now rejected as uncalled-for and groundless. Yet more untenable, and wanting even a semblance of foundation, is the derivation of Latin from Greek, a favourite dogma of times not long past, but at present abandoned by every comparative philologist whose opinion is of the slightest value.

In the Greek language, we reach an antiquity in the recorded history of Indo-European speech considerably higher than we have anywhere else attained. The exact date of its earliest monuments, the grand and unrivalled poems of Homer, the Iliad and Odyssey, cannot, it is true, be determined; but they go back, doubtless, to near the beginning of the thousand years before Christ's birth. From the different parts of Greece, too, as of Italy, we have received records of dialects that subsisted side by side through all the earlier periods of the country's history, until at length (about B.C. 300) the political importance and superior literature of Athens made her idiom, the later Attic, the common language of cultivated Greeks everywhere. The earlier Attic is found first in the writings of the great dramatists, beginning about five centuries before Christ: it is more nearly akin with the earlier Ionic of Homer and Hesiod (before 700 B.C.), and the later Ionic of Herodotus (about 400 B.C.), than with the Doric of Alcman, Pindar, and Theocritus (600-250 B.C.), or the Æolic of Alcæus and Sappho (about 600 B.C.). The differences of the Greek dialects are quite insignificant as compared with those of the Italic, yet they are of no small service to the historical student of the Greek language, since each brings to his knowledge some elements less corrupted and modernized than are to be found in the others, or in the later common tongue.

The modern Greek has also its dialects, respecting which little is known in detail; and it has, besides, its common tongue, the Romaic (as it is ordinarily styled), spoken and written by all the educated Greeks of the present day. This Romaic is very much less altered from the ancient classic language, as spoken by Plato and Demosthenes, than are the modern Romanic languages from the speech of Virgil and Cicero. The difference of the two is even so slight that a party in Greece are now engaged in making the somewhat pedantic and utopian effort to eliminate it altogether, to make the turbulent population of the present petty and insignificant kingdom talk and write as did their heroic forefathers, when, though feeble in numbers, they were the foremost community of the world. Small result is to be looked for from this experiment; should it prove successful, it will be the first time that such a thing has been accomplished in all the history of language.

Of the Asiatic branches of our family, the one which lies nearest us, the Iranian, or Persian, may first engage our attention. Its oldest monuments of well-determined date are the inscriptions—cut on the surface of immense walls of living rock, in the so-called cuneiform characters—by which the Achæmenidan sovereigns of Persia, Darius, Xerxes, and their successors, made imperishable record for posterity of their names and deeds. Fifty years ago, these inscriptions were an unsolved and apparently insoluble enigma; now, by a miracle of human ingenuity and patience, not without the aid of a combination of favouring circumstances wholly impossible at any earlier period, almost every word and every character is fully laid open to our comprehension, and they have been made to yield results of great value both to linguistic and to national history. The oldest of them come from a time about five centuries before Christ, and their extent is sufficient to give us a very distinct idea of the language of those Persians against whom the Greeks so long fought, first for independence, then for empire.

Of about the same age, and even, probably, in part considerably older, are the sacred Scriptures of the religion established by Zoroaster (in his own tongue, Zarathustra)—the book called the Avesta, or Zend-Avesta. The dialect in which these writings are composed goes usually by the name of the Zend; it is also styled the Avestan, and sometimes the Old Bactrian, from the country Bactria, the north-eastern-most region of the great Iranian territory, which is supposed to have been its specific locality. They have been preserved to us by the Parsis of western India, who fled thither from their native country after its reduction under Mohammedan vassalage in the seventh century of our era, and who have ever since faithfully maintained, under Hindu and British protection, the rites of the Magian faith, the pure worship of Ormuzd (Ahura-Mazda, 'the mighty spirit') through the symbol of fire. The Avesta shows two dialects, a younger and an older; some of its hymns and prayers possibly go back to the time of Zoroaster himself—whatever that may have been: it was doubtless more than a thousand years, at least, before Christ—but the bulk of the work is considerably later. Accompanying the Avesta is a version of it, made for the use of the priests, in another and much more modern Iranian dialect, the Pehlevi or Huzvaresh, supposed to have been the literary language of the westernmost provinces of Iran at a period some centuries later than the Christian era, and much mixed with materials derived from the Semitic tongues lying next westward, across the border. A few inscriptions and legends of coins, of the early Sassanian monarchs (after A.D. 226), furnish further specimens of the same or a nearly kindred dialect.

The general body of religious literature belonging to the Parsis of India contains tolerably copious documents of a somewhat younger and much purer Iranian dialect, usually styled the Parsi (sometimes also the Pazend). It comes, without much question, from a more eastern locality than the Pehlevi, and from a time nearly approaching that of the Mohammedan conquest. Finally, after the conquest, and when Persia was thoroughly made over into a province of the Moslem empire, arises, in the tenth century, the modern Persian, and becomes during several centuries, and even to our own day, the vehicle of an abundant and admirable literature, rich in every department, in poetry, fiction, history, philosophy, science. Its first great work, and almost or quite the greatest it has to offer us, is the Shah-Nameh, 'Book of Kings,' of Firdusi (ob. 1020), a true national epic, grand in extent, noble in style, varied in contents, in which is summed up and related at length the history of the land, traditional, legendary, and mythological, as it lay in the minds of the generation by whom was revived the ancient independence and glory of the Persian nationality. For the impoverishment of its grammar by the loss of ancient forms, the modern Persian is almost comparable with the English. It is more nearly related to the language of the Achæmenidan inscriptions than to that of the Avesta, although not the lineal descendant and representative of either. In its later literary use, it is greatly disfigured by the unlimited introduction of words from the Arabic vocabulary.

There are several other language, in regions bordering on or included within the Iranian territory, which stand in such relations with those we have been describing as to be ranked in the same class, although their Iranian attributes are greatly obscured by the changes which have passed upon them since their separation from the principal stock. By far the most important of these is the Armenian, with an abundant literature going back to the fifth century, the era of the Christianization of the Armenian people. Others are the Ossetic, in the Caucasus; the Kurdish, the dialect of the wild mountaineers of the border lands between Persia, Turkey, and Russia; and the Afghan or Pushto, which in very recent times has enjoyed a certain degree of literary cultivation.

We come, finally, to that member of our family which has lived its life within the borders of India. Not all the numerous dialects which fill this immense peninsula, between the impassable wall of the Himalayas and the Indian ocean, own kindred with the Indo-European tongues, but only those of its northern portion, of Hindustan proper, ranging from the Indus to the mouths of the Ganges, together with a certain extent of the sea-coast and its neighbourhood stretching southward on either side. The central mountainous region and the table-lands of the Dekhan yet belong to the aboriginal tribes, who in the north were crowded out or subjugated, at a period lying only just beyond the ken of recorded history, by the Indo-European races, as they intruded themselves through the avenue, the passes on the north-western frontier, by which the conquerors of India have in all ages found entrance. The principal modern dialects of our kindred are the Hindi, Bengali, and Mahratta, each with various subdivisions, and each with a literature of its own, running back only a few centuries. The Hindustani, or Urdu, is a form of the Hindi which grew up in the camps (ûrdû) of the Mohammedan conquerors of India, since the eleventh century, as medium of communication between them and the subject population of central Hindustan, more corrupted in form, and filled with Persian and Arabic words—being thus, as it were, the English of India: it has enjoyed more literary cultivation than any other of the recent dialects, and is the lingua franca, the official language and means of general intercourse, throughout nearly the whole peninsula. The tongue of the roving Gypsies all over Europe, though everywhere strongly tinged with the local idiom of the region of their wanderings, is in its main structure and material a modern Hindu patois: the Gypsies are exiles from India.

Next older than the languages we have mentioned are the Prakrit and the Pali, represented by a literature and inscriptions which come to us in part from before the Christian era. The Pali is the sacred language of the Buddhist religion in the countries lying eastward and south-eastward from India. The Prakrit dialects are chiefly preserved in the Sanskrit dramas, where the unlearned characters, the women, servants, and the like, talk Prakrit—just as, in a modern German theatre, one may hear the lower personages talk the dialects of their own districts, while the higher employ the literary German, the common speech of the educated throughout the country.

The virtual mother of all these dialects is the Sanskrit. For the last twenty-five centuries, at least, the Sanskrit has been no longer a proper vernacular language, but kept artificially in life, as the sacred dialect of Brahmanism and the cultivated tongue of literature and learning; thus occupying a position closely analogous with that held by the Latin since the decline of the western empire, as the language of Roman Catholicism, and the means of communication among the learned of all Europe. It is still taught in the schools of the Brahmanic priesthood, used in the ceremonies of their religion, and spoken and written by their foremost scholars—although, like the Latin in more recent times, much shaken in its sway by the uprise of the modern cultivated dialects, and the decadence of the religion with whose uses it is identified. We possess it in two somewhat varying forms, the classical Sanskrit, and the older idiom of the so-called Vedas, the Bible of the Hindu faith. The former is more altered, by elaborate and long-continued literary and grammatical training, from the condition of a true vernacular, than is almost any other known literary language. Partly for this reason, and partly because, at the time of its establishment and fixation as the learned tongue of all Aryan India, it must have been one among a number of somewhat differing local varieties of Aryan speech, whose differences form a part of the discordance of the later dialects, I have called it above rather their virtual than their actual progenitor: it represents very closely the primitive stock out of which they have all grown, by varying internal development, and by varying influence and admixture of foreign tongues. When and where it was at first a spoken dialect, is out of our power to determine; but it cannot well be regarded as of less age than the earliest Greek records; and it is probably older by centuries. It possesses a most abundant literature, in nearly every department save history; its religious and ethical poetry, its epics, its lyric flights, its dramas, its systems of philosophy and grammar, have been found worthy of high admiration and of profound study by Western scholars; they have even been ranked by some, though very unjustly, as superior to the masterpieces of the Greek and Latin literatures. To fix the chronology of its separate works is a task of the extremest difficulty; but some of them, even in their present form, and the substance of many others, certainly come from a time considerably anterior to the Christian era.

The Vedic dialect is yet more ancient; the earliest portions of the oldest collection, the Rig-Veda ('Veda of hymns'), must, it is believed, date from nearly or quite two thousand years before Christ. The considerations from which this age is deduced for them are of a general and inexact character, yet tolerably clear in their indications. Thus, for example, the hymns of the Vedas were chiefly composed on the banks of the Indus and its tributaries, when the great valley of the Ganges was as yet unknown to the Aryan immigrants; and they present the elephant as still a wondered-at and little-known animal: while the earliest tidings of India which we have from without show us great kingdoms on the Ganges, and the elephant reduced to the service of man, both in war and in peace. Buddhism, too, which is well known to have preceded by several centuries the birth of Christ, was a revolt against the oppressive domination of the Brahmanic hierarchy; and in the Vedas are to be seen only the germs of Brahmanism, not yet developed: no hierarchy, no system of castes, no vestige of the doctrine of transmigration. The conclusions drawn from a study of the internal history and connection of the different classes of works composing the sacred literature of India—which follow one another, in a close succession of expositions, rules, and comments, from a time not much later than that of the more recent hymns down to the historical period—point also to the same age. The Vedas are thus by not less than a thousand years the earliest documents for the history of Indo-European language—for the history, moreover, of Indo-European conditions and institutions. The civil constitution, the religious rites, the mythologic fancies, the manners and customs, which they depict, have a peculiarly original and primitive aspect, seeming to exhibit a far nearer likeness to what once belonged to the whole Indo-European family than is anywhere else to be attained. The Vedas appear rather like an Indo-European than an Indian record; they are the property rather of the whole family than of a single branch.

Much of the same character appertains to the classical Sanskrit: it is both earlier in chronologic period and more primitive in internal character than any other language of the whole great family. Its peculiar value lies in its special conservation of primitive material and forms, in the transparency of its structure, in its degree of freedom from the corrupting and disguising effects of phonetic change, from obliteration of original meaning and application. We must beware of supposing that at all points, in every item of structure, it is the superior of the other Indo-European tongues, or that it constitutes an infallible norm by which their material is to be judged; on the contrary, each of the other branches here and there excels it, offering some remains of early Indo-European speech which it has lost; but to it must be freely conceded the merit of having retained, out of the common stock, more than any one among them, almost more than they all. Exaggerated and unfounded claims are often put forward in its behalf by those who do not fully understand the true sources of its value: its alphabet, though rich and very harmoniously developed, does not cover more than about two-thirds of our English system of spoken sounds; as an instrument of the expression of thought it has very serious and conspicuous defects, being inferior—especially in its handling of the verb (the soul of the sentence), in a loose and bald syntactical arrangement, and in an excessive use of compounds—not only to the Greek, but to almost every other cultivated Indo-European tongue; nor (as has been already hinted) can its literature sustain a moment's comparison with those of the classical languages. It is to be prized chiefly as a historical document, casting inestimable light upon the earliest development of the common speech of the Indo-European family, and the relations of its members. Had all its literature besides perished, leaving us only a grammar of its forms and a dictionary of its material, it would still in a great measure retain this character; were but a fragment of one of its texts saved, as has been the case with the Mœso-Gothic, it would still vindicate its right to a place at the head of all the languages of the family. It may easily be appreciated, then, what an impulse to the historical study of language, then just struggling into existence by the comparison of the tongues of Europe, was given by the discovery and investigation of this new dialect, having a structure that so invited and facilitated historic analysis, and even presented by the native grammatical science in an analyzed condition, with roots, themes, and affixes carefully separated, distinctly catalogued, and defined in meaning and office. In all researches into the beginnings of Indo-European speech, the genesis of roots and forms, its assistance is indispensable, and its authority of greatest weight. It often has been and still is wrongly estimated and misapplied by incautious or ill-instructed investigators; it is sometimes treated as if it were the mother of the Indo-European dialects, as the Latin of the modern Romanic tongues, instead of merely their eldest sister, like the Mœso-Gothic among the Germanic languages; it is unduly brought in to aid the inter-comparison of dialects of a single branch, and its peculiar developments, its special laws of euphony or construction, are sought to be forced upon them; the facts it presents are erroneously accepted as ultimate, cutting off further inquiry; portions of its existing material which are of modern growth, or the artificial productions of Hindu scholasticism, are perversely used as of avail for Indo-European etymology: and such abuse has naturally provoked from some scholars a distrust of its genuine claims to regard: but, stripping off all exaggerations, and making all due allowances, the Sanskrit is still the main-stay of Indo-European philology; it gave the science a rapid development which nothing else could have given; it imparted to its conclusions a fulness and certainty which would have been otherwise unattainable.

Such is the constitution of the grand division of human speech to which our own language belongs. That its limits have been everywhere traced with entire exactness cannot, of course, be claimed; other existing dialects may yet make good their claim to be included in it—and it is beyond all reasonable question that, as many of its sub-branches have perished without leaving a record, so various of its branches, fully coördinate with those we have reviewed, must have met a like fate. We may now proceed to glance briefly at some of the grounds of the preëminent importance with which it is invested.

One source of the special interest which we feel in the study of Indo-European language lies in the fact that our own tongue is one of its branches. In the moral and intellectual world, not less than in the physical, everything cannot but appear larger in our eyes according as it is nearer to us. This would be a valid consideration with any race upon earth, since, for each, its own means of communication and instrument of thought is also the record of its past history, and must be its agency of future improvement in culture, and therefore calls for more study in order to its fuller comprehension, and its development and elevation, than should be given to any other tongue, of however superior intrinsic value. But we are further justified in our somewhat exclusive interest by the position which our languages, and the races which speak them, hold among other languages and races. It is true, as was claimed at the outset of these lectures, that linguistic science, as a branch of human history, aims at universality, and finds the tongues of the humblest tribes as essential to her completeness as those of the most cultivated and gifted nations; but it is also true that, mindful of proportion, she passes more lightly over the one, to give her longer and more engrossed attention to the other. While the weal and woe of every individual that ever lived goes to make up the sum of human interests, with which our human nature both justifies and demands our sympathy, we cannot but linger longest and with keenest participation over the fortunes of those who have played a great part among their fellows, whose deeds and words have had a wide and deep-reaching influence. And this is, in a very marked degree, the character of the Indo-European race. Its first entrance as an actor into what we are accustomed to call universal history, or that drama of action and influence whose denouement is the culture of the modern European nations, was in the far East, in the Persian empire of Cyrus and his successors. This founded itself upon the ruins and relics of more ancient empires and cultures, belonging to other peoples, in part Semitic, in part of obscurer kindred. For the Indo-Europeans were, of all the great civilizing and governing races, the last to commence their career. Not only in Mesopotamia, but also in Egypt and China, the light of knowledge burned brightly, and great deeds were done, whereof the world will never lose the memory, while the tribes of our kindred were wandering savages, or weak and insignificant communities, struggling for existence. The Persian empire, in its conquering march westward, was first checked by one of these humble communities, the little jarring confederation of Greek states and cities, destined to become, notwithstanding its scanty numbers, the real founder of Indo-European preëminence. Greece, enriching itself with elements drawn from the decaying institutions of older races, assimilated them, and made them lively and life-giving, with an energy of genius unrivalled elsewhere in the annals of the world. The wider the range of our historical study, the more are we penetrated with the transcendent ability of the Greek race. In art, literature, and science, it has been what the Hebrew race has been in religion, and its influence has been hardly less unlimited, in space and in time.

It seemed at one period, as is well known, that Greece would succeed to the imperial throne of Persia, subjecting the civilized world to her sway; but the prospect lasted but for a moment: the sceptre of universal dominion slipped from the hands of Alexander's successors, and soon passed over into the keeping of another and younger branch of the same family. Rome, appropriating the fruits of Greek culture, and adding an organizing and assimilating force peculiarly her own, went forth to give laws to all nations, and to impose upon them a unity of civilization and of social and political institutions. And if Christianity was of Semitic birth, Greeks and Romans gave it universality. Rejected by the race which should have especially cherished it, it was taken up and propagated by the Indo-Europeans, and added a new unity, a religious one, to the forces by which Rome bound together the interests and fates of mankind.

Now came the turn of yet another branch, the Germanic. This had, indeed, only the subordinate part to play of aiding in the downfall of the old order of things, and preparing the way for a new and more vigorous growth. Its tribes ravaged Europe from east to west, and even to the farthest southern coasts, giving ruling class and monarch to nearly every country of the continent. But centuries of weakness and confusion were the first result of this great up-turning, and it even appeared for a time as if the dominion of the world were destined to be usurped by another race. The Semites, inspired with the furious zeal of a new religion, Mohammedanism, broke from their deserts and overran the fairest parts of Asia and Africa; and their conquering hosts entered Europe at either extremity, establishing themselves firmly, and pushing forward to take possession of the rest. They recoiled, at last, before the reviving might of the superior race, and the last and grandest era of Indo-European supremacy began, the era in the midst of which we now live. For the past few centuries, the European nations have stood foremost, without a rival, in the world's history. They are the enlightened and the enlighteners of mankind. They alone are extending the sphere of human knowledge, investigating the nature of matter and of mind, and tracing out their exhibition in the past history and present condition of the earth and its inhabitants. They alone have a surplus stock of intelligent energy, which is constantly pushing beyond its old boundaries, and spurns all limit to its action. The network of their activity embraces the globe; their ships are in every sea between the poles, for exploration, for trade, or for conquest; the weaker races are learning their civilization, falling under their authority, or perishing off the face of the land, from inherent inability to stand before them. They have appropriated, and converted into outlying provinces of their race and culture, the twin world of the West, and the insular continent of the south-eastern seas, while their lesser colonies dot the whole surface of the inhabitable globe. They have inherited from its ancient possessors the sceptre of universal dominion, over a world vastly enlarged beyond that to which were limited the knowledge and the power of former times: and they are worthy to wield it, since their sway brings, upon the whole, physical well-being, knowledge, morality, and religion to those over whom it is extended.

All that speciality of interest, then, which cleaves to historical investigations respecting the origin, the earliest conditions, the migrations, the mutual intercourse and influence, and the intercourse with outside races, of that division of mankind which has shown itself as the most gifted, as possessing the highest character and fulfilling the noblest destiny, among all who have peopled the earth since the first dawn of time, belongs, of right and of necessity, to Indo-European philology.

It may, indeed, be urged that this is an interest lying somewhat apart from the strict domain of linguistic science, whose prime concern is with speech itself, not with the characters or acts of those who speak. Yet, as was pointed out in our first lecture, the study of language is not introspective merely; they would unduly narrow its sphere and restrict its scope who should limit it to the examination of linguistic facts: these are so inextricably intertwined with historical facts, so dependent upon and developed out of them, that the two cannot be separated in consideration and treatment; one chief department of the value of the science lies in its capacity to throw light upon the history of human races. The importance of the Indo-European races in history is, then, legitimately to be included among the titles of Indo-European philology to the first attention of the linguistic scholar. Moreover, since the relation between the capacity of a race and the character of the tongue originated and elaborated by that race is a direct and necessary one, it could not but be the case that the speech of the most eminently and harmoniously endowed part of mankind should itself be of highest character and most harmonious development, and so the most worthy object of study, in its structure and its relations to mind and thought. And this advantage also, as we shall see more plainly hereafter, is in fact found to belong to Indo-European language: in the classification of all human speech it takes, unchallenged, the foremost rank.

But these considerations, weighty as they are, do not fully explain the specially intimate bond subsisting between general linguistic science and the study of Indo-European speech. Not only did the establishment of the unity of that family, and the determination of the relations of its members, constitute the most brilliant achievement of the new science; they were also its foundation; it began with the recognition of these truths, and has developed with their elaboration. The reason is not difficult to discover: Indo-European language alone furnished such a grand body of related facts as the science needed for a sure basis. Its dialects have a range, in the variety of their forms and in the length of the period of development covered by them, which is sought elsewhere in vain. They illustrate the processes of linguistic growth upon an unrivalled scale, and from a primitive era to which we can make but an imperfect approach among the other languages of mankind. Portions of the Chinese literature, it is true, are nearly or quite as old as anything Indo-European, and the Chinese language, as will be shown later, is in some respects more primitive in its structure than any other human tongue; but what it was at the beginning, that it has ever since remained, a solitary example of a language almost destitute of a history. Egypt has records to show of an age surpassing that of any other known monuments of human speech; but they are of scanty and enigmatical content, and the Egyptian tongue also stands comparatively alone, without descendants, and almost without relatives. The Semitic languages come nearest to offering a worthy parallel; but they, too, fall far short of it. The earliest Hebrew documents are not greatly exceeded in antiquity by any others, and the Hebrew with its related dialects, ancient and modern, fills up a linguistic scheme of no small wealth; yet Semitic variety is, after all, but poor and scanty as compared with Indo-European; Semitic language possesses a toughness and rigidity of structure which has made its history vastly less full of instructive change; and its beginnings are of unsurpassed obscurity. The Semitic languages are rather a group of closely kindred dialects than a family of widely varied branches: their whole yield to linguistic science is hardly more than might be won from a single subdivision of Indo-European speech, like the Germanic or Romanic. None of the other great races into which mankind is divided cover with their dialects, to any noteworthy extent, time as well as space; for the most part, we know nothing more respecting their speech than is to be read in its present living forms. Now it is so obvious as hardly to require to be pointed out, that a science whose method is prevailingly historical, which seeks to arrive at an understanding of the nature, office, and source of language by studying its gradual growth, by tracing out the changes it has undergone in passing from generation to generation, from race to race, must depend for the soundness of its methods and the sureness of its results upon the fulness of illustration of these historical changes furnished by the material of its investigations. It is true that the student's historical researches are not wholly baffled by the absence of older dialects, with whose forms he may compare those of more modern date. Something of the development of every language is indicated in its own structure with sufficient clearness to be read by analytic study. Yet more is to be traced out by means of the comparison of kindred contemporaneous dialects; for, in their descent from their common ancestor, it can hardly be that each one will not have preserved some portion of the primitive material which the others have lost. Thus—to illustrate briefly by reference to one or two of our former examples—the identity of our suffix ly, in such words as godly and truly, with the adjective like might perhaps have been conjectured from the English alone; and it is made virtually certain by comparison with the modern German (göttlich, treulich) or Netherlandish (goddelijk, waarlijk); it does not absolutely need a reference to older dialects, like the Anglo-Saxon or Gothic, for its establishment. Again, not only the Sanskrit and other ancient languages exhibit the full form asmi, whence comes our I am, but the same is also to be found almost unaltered in the present Lithuanian esmi. But, even if philological skill and acumen had led the student of Germanic language to the conjecture that I loved is originally I love-did, it must ever have remained a conjecture only, a mere plausible hypothesis, but for the accident which caused the preservation to our day of the fragment of manuscript containing a part of Bishop Ulfilas's Gothic Bible. And a host of points in the structure of the tongues of our Germanic branch which still remain obscure would, as we know, be cleared up, had we in our possession relics of them at a yet earlier stage of their separate growth. The extent to which the history of a body of languages may be penetrated by the comparison of contemporary dialects alone will, of course, vary greatly in different cases; depending, in the first place, upon the number, variety, and degree of relation of the dialects, and, in the second place, upon their joint and several measure of conservation of ancient forms: but it is also evident that the results thus arrived at for modern tongues will be, upon the whole, both scanty and dubious, compared with those obtained by comparing them with ancient dialects of the same stock. Occasionally, within the narrow limits of a single branch or group, the student enjoys the advantage of access to the parent tongue itself, from which the more recent idioms are almost bodily derived: thus, for example, our possession of the Latin gives to our readings of the history of the Romanic tongues, our determination of the laws which have governed their growth, a vastly higher degree of definiteness and certainty than we could reach if we only knew that such a parent tongue must have existed, and had to restore its forms by careful comparison and deduction. Next in value to this is the advantage of commanding a rich body of older and younger dialects of the same lineage, wherein the common speech is beheld at nearer and remoter distances from its source, so that we can discover the direction of its currents, and fill out with less of uncertainty those parts of their network of which the record is obliterated. This secondary advantage we enjoy in the Germanic, the Persian, the Indian branches of Indo-European speech; and, among the grander divisions of human language, we enjoy it to an extent elsewhere unapproached in the Indo-European family, that immense and varied body of allied forms of speech, whose lines of historic development are seen to cover a period of between three and four thousand years, as they converge toward a meeting in a yet remoter past.

Herein lies the sufficient explanation of that intimate connection, that almost coincidence, which we have noticed between the development of Indo-European comparative philology and that of the general science of language. In order to comprehend human language in every part, the student would wish to have its whole growth, in all its divisions and subdivisions; through all its phases, laid before him for inspection in full authentic documents. Since, however, anything like this is impossible, he has done the best that lay within his power: he has thrown himself into that department of speech which had the largest share of its history thus illustrated, and by studying that has tried to learn how to deal with the yet more scanty and fragmentary materials presented him in other departments. Here could be formed the desired nucleus of a science; here the general laws of linguistic life could be discovered; here could be worked out those methods and processes which, with such modifications as the varying circumstances rendered necessary, should be applied in the investigation of other types of language also. The foundation was broad enough to build up a shapely and many-sided edifice upon. Yet the study of Indo-European language is not the science of language. Such is the diversity in unity of human speech that exclusive attention to any one of its types could only give us partial and false views of its nature and history. Endlessly as the dialects of our family appear to differ from one another, they have a distinct common character, which is brought to our apprehension only when we compare them with those of other stock; they are far from exhausting the variety of expression which the human mind is capable of devising for its thought; the linguist who trains himself in them alone will be liable to narrowness of vision, and will stumble when he comes to walk in other fields. We claim only that their inner character and outer circumstances combine to give them the first place in the regard of the linguistic scholar; that their investigation will constitute in the future, as it has done in the past, a chief object of his study; and that their complete elucidation is both the most attainable and the most desirable and rewarding object proposed to itself by linguistic science.

The general method of linguistic research has already been variously set forth and illustrated, in an incidental way; but a summary recapitulation of its principles, with fuller reference to the grounds on which they are founded, will not be amiss at this point in our progress. The end sought by the scientific investigator of language, it will be remembered, is not a mere apprehension and exposition, however full and systematic, of the phenomena of a language, or of all human speech—of its words, its forms, its rules, its usages: that is work for grammarians and lexicographers. He strives to discover the why of everything: why these words, these affixes, have such and such meanings; why usage is thus, and not otherwise; why so many and such words and forms, and they only, are found in a given tongue—and so on, in ever farther-reaching inquiry, back even to the question, why we speak at all. And since it appears that every existing or recorded dialect, and every word composing it, is the altered successor, altered in both form and meaning, of some other and earlier one; since all known language has been made what it is, out of something more original, by action proceeding from the minds of those who have used it, its examination must be conducted historically, like that of any other institution which has had a historic growth and development. All human speech has been during long ages modified, was even perhaps in the first place produced, by human capacities, as impelled by human necessities and governed by human circumstances; it has become what these influences by their gradual action have made it: it, on the one hand, is to be understood only as their product; they, on the other hand, are to be read in the effects which they have wrought upon it. To trace out the transformations of language, following it backward through its successive stages even to its very beginnings, if we can reach so far; to infer from the changes which it is undergoing and has undergone the nature and way of action of the forces which govern it; from these and from the observed character of its beginnings to arrive at a comprehension of its origin—such are the inquiries which occupy the attention of the linguistic scholar, and which must guide him to his ultimate conclusions respecting the nature of speech as an instrumentality of communication and of thought, and its value as a means of human progress.

And as in its general character, so also in its details, the process of investigation is historical. We have already seen (lecture second, p. 54) that the whole structure of our science rests upon the study of individual words; the labours of the etymologist must precede and prepare the way for everything that is to follow. But every etymological question is strictly a historical one; it concerns the steps of a historical process, as shown by historical evidences; it implies a judgment of the value of testimony, and a recognition of the truth fairly deducible therefrom. What is proved respecting the origin and changes of each particular word by all the evidence within reach, is the etymologist's ever-recurring inquiry. To answer it successfully he needs a combination of many qualities; he must be, in fact, a whole court in himself: the acuteness, perseverance, and enterprise of the advocate must be his, to gather every particle of testimony, every analogy, every decision, bearing upon the case in hand; he must play the part of the opposing counsel, in carefully sifting the collected evidence, testing the character and disinterestedness of the witnesses, cross-examining them to expose their blunders and inconsistencies; he must have, above all, the learning and candour of the judge, that he may sum up and give judgment impartially, neither denying the right which is fairly established, nor allowing that which rests on uncertain allegation and insufficient proof. In short, the same gifts and habits of mind which make the successful historian of events are wanted also to make the successful historian of words.

The ill-repute in which etymology and those who follow it are held in common opinion is a telling indication of the difficulty attending its practice. The uncertainty and arbitrariness of its prevailing methods, the absurdity of its results, have been the theme of many a cutting and well-directed gibe. It has in all ages been a tempting occupation to curious minds, and always a slippery one. An incalculable amount of human ingenuity has been wasted in its false pursuit. Men eminent for acuteness and sound judgment in other departments of intellectual labour have in this been guilty of folly unaccountable. It has been often remarked that the Greeks and Romans, when once engaged in an etymological inquiry, seem to have taken leave of their common sense. Great as were the advantages offered by the Sanskrit language to its native analysts, in the regularity of its structure and the small proportion of obscure words which it contained, they stumbled continually as soon as they left the plain track of the commonest and clearest derivations, and their religious, philosophical, and grammatical books are filled with word-genealogies as fanciful and unsound as those of the classic writers. In no one respect does the linguistic science of the present day show its radical superiority to that of former times more clearly than in the style and method of its etymologies: upon these, indeed, is its superiority directly founded.

The grand means, now, of modern etymological research is the extensive comparison of kindred forms. How this should be so appears clearly enough from what has been already taught respecting the growth of dialects and the genetical connections of languages. If spoken tongues stood apart from one another, each a separate and isolated entity, they would afford no scope for the comparative method. As such entities the ancient philology regarded them; or, if their relationship was in some cases recognized, it was wrongly apprehended and perversely applied—as when, for instance, the Latin was looked upon as derived from the Greek, and its words were sought to be etymologized out of the Greek lexicon, as corrupted forms of Greek vocables. In the view of the present science, while each existing dialect is the descendant of an older tongue, so other existing dialects are equally descendants of the same tongue. All have kept a part, and lost a part, of the material of their common inheritance; all have preserved portions of it in a comparatively unchanged form, while they have altered other portions perhaps past recognition. But, while thus agreed in the general fact and the general methods of change, they differ indefinitely from one another in the details of the changes effected. Each has saved something which others have lost, or kept in pristine purity what they have obscured or overlaid: or else, from their variously modified forms can be deduced with confidence the original whence these severally diverged. Every word, then, in whose examination the linguistic scholar engages, is to be first set alongside its correspondents or analogues in other related languages, that its history may be read aright. Thus the deficiencies of the evidence which each member of a connected group of dialects contains respecting its own genesis and growth are made up, in greater or less degree, by the rest, and historical results are reached having a greatly increased fulness and certainty. The establishment of a grand family of related languages, like the Indo-European, makes each member contribute, either immediately or mediately, to the elucidation of every other.

The great prominence in the new science of language of this comparative method gave that science its familiar title of "comparative philology," a title which is not yet lost in popular usage, although now fully outgrown and antiquated. It designated very suitably the early growing phase of linguistic study, that of the gathering and sifting of material, the elaboration of methods, the establishment of rules, the deduction of first general results; it still properly designates the process by which the study is extended and perfected; but to call the whole science any longer "comparative philology" is not less inappropriate than to call the science of zoölogy "comparative anatomy," or botanical science the "comparison of plants."

But the comparative method, as we must not fail to notice, is no security against loose and false etymologizing; it is not less liable to abuse than any other good thing. If it is to be made fruitful of results for the advancement of science, it must not be wielded arbitrarily and wildly; it must have its fixed rules of application. Some appear to imagine that, in order to earn the title of "comparative philologist," they have but to take some given language and run with it into all the ends of the earth, collating its material and forms with those of any other tongue they may please to select. But that which makes the value of comparison—namely, genetical relationship—also determines the way in which it shall be rendered valuable. We compare in order to bring to light resemblances which have their ground and explanation in a real historical identity of origin. We must proceed, then, as in any other genealogical inquiry, by tracing the different lines of descent backward from step to step toward their points of convergence. The work of comparison is begun between the tongues most nearly related, and is gradually extended to those whose connection is more and more remote. We first set up, for example, a group like the Germanic, and by the study of its internal relations learn to comprehend its latest history, distinguishing and setting apart all that is the result of independent growth and change among its dialects, recognizing what in it is original, and therefore fair subject of comparison with the results of a like process performed upon the other branches of the same family. It needs not, indeed that the restoration of primitive Germanic speech should be made complete before any farther step is taken; there are correspondences so conspicuous and palpable running through all the varieties of Indo-European speech, that, the unity of the family having been once established, they are at a glance seen and accepted at their true value. But only a small part of the analogies of two more distantly related languages are of this character, and their recognition will be made both complete and trustworthy in proportion as the nearer congeners of each language are first subjected to comparison. If English were the only existing Germanic tongue, we could still compare it with Attic Greek, and point out a host of coincidences which would prove their common origin; but, as things are, to conduct our investigation in this way, leaving out of sight the related dialects on each side, would be most unsound and unphilological; it would render us liable to waste no small share of our effort upon those parts of English which are peculiar, of latest growth, and can have no genetic connection whatever with aught in the Greek: it would expose us, on the one hand, to make false identifications (as between our whole and the Greek holos, 'entire'); and, on the other hand, to find diversity where the help of older dialectic forms on both sides would show striking resemblance. What analogy, for instance, do we discern between our bear, in they bear, and Greek pherousi? but comparison of the other Germanic dialects allows us to trace bear directly back to a Germanic form berand, and Doric Greek gives us pheronti, from which comes pherousi by one of the regular euphonic rules of the language; the law of permutation of mutes in the Germanic languages (see above, p. 97) exhibits b as the regular correspondent in Low German dialects to the original aspirate ph; and the historical identity of the two words compared, in root and termination, is thus put beyond the reach of cavil.

Yet more contrary to sound method would it be, for example, to compare directly English, Portuguese, Persian and Bengali, four of the latest and most altered representatives of the four great branches of Indo-European speech to which they severally belong. Nothing, or almost nothing, that is peculiar to the Bengali as compared with the Sanskrit, to the Persian as compared with the ancient Avestan and Achæmenidan dialects, to the Portuguese as compared with the Latin, can be historically connected with what belongs to English or any other Germanic tongue. Their ties of mutual relationship run backward through those older representatives of the branches, and are to be sought and traced there.

But worst of all is the drawing out of alleged correspondences, and the fabrication of etymologies, between such languages as the English—or, indeed, any Indo-European dialect—on the ono hand, and the Hebrew, or the Finnish, or the Chinese, on the other. Each of these last is the fully recognized member of a well-established family of languages, distinct from the Indo-European. If there be genetic relation between either of them and an Indo-European language, it must lie back of the whole grammatical development of their respective families, and can only be brought to light by the reduction of each, though means of the most penetrating and exhaustive study of the dialects confessedly akin with it, to its primitive form, as cleared of all the growth and change wrought upon it by ages of separation. There may be scores, or hundreds, of apparent resemblances between them, but these are worthless as signs of relationship until an investigation not less profound than we have indicated shall show that they are not merely superficial and delusive.

Let it not be supposed that we are reasoning in a vicious circle, in thus requiring that two languages shall have been proved related before the correspondences which are to show their relationship shall be accepted as real. We are only setting forth the essentially cumulative nature of the evidences of linguistic connection. The first processes of comparison by which it is sought to establish the position and relations of a new language are tentative merely. No sound linguist is unmindful of the two opposing possibilities which interfere with the certainty of his conclusions: first, that seeming coincidences may turn out accidental and illusory only; second, that beneath apparent discordance may be hidden genetic identity. With every new analogy which his researches bring to view, his confidence in the genuineness and historic value of those already found is increased. And when, examining each separate fact in all the light that he can cast upon it, from sources near and distant, he has at length fully satisfied himself that two tongues are fundamentally related, their whole mutual aspect is thereby modified; he becomes expectant of signs of relationship everywhere, and looks for them in phenomena which would not otherwise attract his attention for a moment. When, on the contrary, an orderly and thorough examination, proceeding from the nearer to the remoter degrees of connection, has demonstrated the position of two languages in two diverse families, the weight of historic probability is shifted to the other scale, and makes directly against the interpretation of their surface resemblances as the effect of anything but accident or borrowing.

The new etymological science differs from the old, not in the character of the results which it is willing to admit, but in the character of the evidence on which it is willing to admit them. It will even derive lucus, 'grove,' from non lucendo, 'its not shining there,' if only historical proof of the derivation be furnished. It finds no difficulty in recognizing as identical two words like the French évêque and the English bishop, which have not a single sound or letter in common; for each is readily traceable back to the Greek episkopos.^[3] But it does not draw thence the conclusion that, in this or in any other pair of languages, two words meaning the same thing may, whatever their seeming discordance, be assumed to be one, or are likely to be proved one: it waits for the demonstration in each separate case. The claim made in our third lecture, that, in the history of linguistic changes, any given sound may pass over into any other, any given meaning become modified to its opposite, or to something apparently totally unconnected with it, may seem to take away from etymology all reliable basis; but it is not so; for the same researches which establish this claim show also the difference between those facile changes which may be looked for everywhere, and the exceptional ones which only direct and convincing evidence can force us to accept as actual in any language; they teach us to study the laws of transition of each separate language as part of its idiosyncrasy, and to refrain from applying remote and doubtful analogies in the settlement of difficult questions.

In short, the modern science of language imposes upon all who pursue it thoroughness and caution. It requires that every case be examined in all its bearings. It refuses to accept results not founded on an exhaustive treatment of all the attainable evidence. It furnishes no instruments of research which may not be turned to false uses, and made to yield false results, in careless and unskilful hands. It supplies nothing which can take the place of sound learning and critical judgment. Even those who are most familiar with its methods may make lamentable failures when they come to apply them to a language of which they have only superficial knowledge,^[4] or which they compare directly with some distant tongue, regardless of its relations in its own family, and of its history as determined by comparison with these. A scholar profoundly versed in the comparative philology of some special group of languages, and whom we gladly suffer to instruct us as to their development, may have nothing to say that is worth our listening to, when he would fain trace their remoter connections with groups of which he knows little or nothing. Notwithstanding the immense progress which the study of language has made during the past few years, the world is still full of hasty generalizers, who would rather skim wide and difficult conclusions off the surface of half-examined facts than wait to gather them as the fruits of slow and laborious research. The greater part of the rubbish which is even now heaping up in the path of our science, encumbering its progress, comes from the neglect of these simple principles: that no man is qualified to compare fruitfully two languages or groups who is not deeply grounded in the knowledge of both, and that no language can be fruitfully compared with others which stand, or are presumed to stand, in a more distant relationship with it, until it has been first compared with its own next of kin.

We see, it may be farther remarked, upon how narrow and imperfect a basis those comparative philologists build who are content with a facile setting side by side of words; whose materials are simple vocabularies, longer or shorter, of terms representing common ideas. There was a period in the history of linguistic science when this was the true method of investigation, and it still continues to be useful in certain departments of the field of research. It is the first experimental process; it determines the nearest and most obvious groupings, and prepares the way for more penetrating study. Travellers, explorers, in regions exhibiting great diversity of idiom and destitute of literary records—like our western wilds, or the vast plains of inner Africa—do essential service by gathering and supplying such material, anything better being rendered inaccessible by lack of leisure, opportunity, or practice. But it must be regarded as provisional and introductory, acceptable only because the best that is to be had. Genetic correspondences in limited lists of words, however skilfully selected, are apt to be conspicuous only when the tongues they represent are of near kindred; and even then they may be in no small measure obscured or counterbalanced by discordances, so that deeper and closer study is needed, in order to bring out satisfactorily to view the fact and degree of relationship. Penetration of the secrets of linguistic structure and growth, discovery of correspondences which lie out of the reach of careless and uninstructed eyes, rejection of deceptive resemblances which have no historical foundation—these are the most important part of the linguistic student's work. Surface collation without genetic analysis, as far-reaching as the attainable evidence allows, is but a travesty of the methods of comparative philology.

Another not infrequent misapprehension of etymologic study consists in limiting its sphere of action to a tracing out of the correspondences of words. This is, indeed, as we have called it, the fundamental stage, on the solidity of which depends the security of all the rest of the structure; but it is only that. Comparative etymology, like chemistry, runs into an infinity of detail, in which the mind of the student is sometimes entangled, and his effort engrossed; it has its special rules and methods, which admit within certain limits of being mechanically applied, by one ignorant or heedless of their true ground and meaning. Many a man is a skilful and successful hunter of verbal connections whose views of linguistic science are of the crudest and most imperfect character. Not only does he thus miss what ought to be his highest reward, the recognition of those wide relations and great truths to which his study of words should conduct him, but his whole work lacks its proper basis, and is liable to prove weak at any point. The history of words is inextricably bound up with that of human thought and life and action, and cannot be read without it. We fully understand no word till we comprehend the motives and conditions that called it forth and determined its form. The word money, for example, is not explained when we have marshalled the whole array of its correspondents in all European tongues, and traced them up to their source in the Latin moneta: all the historical circumstances which have caused a term once limited to an obscure city to be current now in the mouths of such immense communities; the wants and devices of civilization and commerce which have created the thing designated by the word and made it what it is; the outward circumstances and mental associations which, by successive changes, have worked out the name from a root signifying 'to think;' the structure of organ, and the habits of utterance—in themselves and in their origin—which have metamorphosed monéta into móney:—all this, and more, is necessary to the linguistic scholar's perfect mastery of this single term. There is no limit to the extent to which the roots of being of almost every word ramify thus through the whole structure of the tongue to which it belongs, or even of many tongues, and through the history of the people who speak them: if we are left in most cases to come far short of the full knowledge which we crave, we at least should not fail to crave it, and to grasp after all of it that lies within our reach.

We have been regarding linguistic comparison as what it primarily and essentially is, the effective means of determining genetical relationship, and investigating the historical development of languages. But we must guard against leaving the impression that languages can be compared for no other purposes than these. In those wide generalizations wherein we regard speech as a human faculty, and its phenomena as illustrating the nature of mind, the processes of thought, the progress of culture, it is often not less important to put side by side that which in spoken language is analogous in office but discordant in origin than that which is accordant in both. The variety of human expression is well-nigh infinite, and no part of it ought to escape the notice of the linguistic student. The comparative method, if only it be begun and carried on aright—if the different objects of the genetic and the analogic comparison be kept steadily in view, and their results not confounded with one another—need not be restricted in its application, until, starting from any centre, it shall have comprehended the whole circle of human speech.

Notes

1. See the works of Marsh, Craik, and others; and especially, for a clear and succinct view of the history and connections of English speech, with grammatical analyses and illustrative specimens, the work of Professor Hadley, already once referred to, on p. 84.
2. In sketching the relations of the Slavonic languages, I follow the authority of Professor August Schleicher, in the Beiträge zur Vergleichenden Sprachforschung, vol. i., p. 1 seq.
3. Evêque, earlier evesque, evesc, represents the syllables episk, while bishop, earlier biskop, represents the syllables piskop. Each has saved, and still accents, the accented syllable of the original; but the French, whose words are prevailingly accented on their final syllables, has dropped off all that followed it; while the Germanic tongues, accenting more usually the penult in words of this structure, has retained the succeeding syllable.
4. Thus, as a striking example and warning, hardly a more utter caricature of the comparative method is to be met with than that given by Bopp, the great founder and author of the method, himself, in the papers in which he attempts to prove the Malay-Polynesian and the Caucasian languages entitled to a place in the Indo-European family.