Lectures on The Science of Language Part 5

_as-si_ _es-si_ _es-si_.

But here Sanskrit, as far back as its history can be traced, has reduced _a.s.si_ to _asi_; and it would be impossible to suppose that the perfect, or, as they are sometimes called, organic, forms in Greek and Lithuanian, _es-si_, could first have pa.s.sed through the mutilated state of the Sanskrit _asi_.

The third person is the same in Sanskrit, Greek, and Lithuanian, _as-ti_ or _es-ti_; and, with the loss of the final _i_, we recognize the Latin _est_, Gothic _ist_, and Russian _est'_.

The same auxiliary verb can be made to furnish sufficient proof that Latin never could have pa.s.sed through the Greek, or what used to be called the Pelasgic stage, but that both are independent modifications of the same original language. In the singular, Latin is less primitive than Greek; for _sum_ stands for _es-um_, _es_ for _es-is_, _est_ for _es-ti_. In the first person plural, too, _sumus_ stands for _es-umus_, the Greek _es-mes_, the Sanskrit _'smas_. The second person _es-tis_, is equal to Greek _es-te_, and more primitive than Sanskrit _stha_. But in the third person plural Latin is more primitive than Greek. The regular form would be _as-anti_; this, in Sanskrit, is changed into _santi_. In Greek, the initial _s_ is dropped, and the aeolic _enti_, is finally reduced to _eisi_. The Latin, on the contrary, has kept the radical _s_, and it would be perfectly impossible to derive the Latin _sunt_ from the Greek _eisi_.

I need hardly say that the modern English, _I am_, _thou art_, _he is_, are only secondary modifications of the same primitive verb. We find in Gothic-

_im_ for _ism_ _is_ for _iss_ _ist_.

The Anglo-Saxon changes the _s_ into _r_, thus giving-

_eom_ for _eorm_, plural _sind_ for _isind_.

_eart_ for _ears_, plural _sind_ _is_ for _ist_, plural _sind_

By applying this test to all languages, the founders of comparative philology soon reduced the princ.i.p.al dialects of Europe and Asia to certain families, and they were able in each family to distinguish different branches, each consisting again of numerous dialects, both ancient and modern.

There are many languages, however, which as yet have not been reduced to families, and though there is no reason to doubt that some of them will hereafter be comprehended in a system of genealogical cla.s.sification, it is right to guard from the beginning against the common, but altogether gratuitous supposition, that the principle of genealogical cla.s.sification must be applicable to all. Genealogical cla.s.sification is no doubt the most perfect of all cla.s.sifications, but there are but few branches of physical science in which it can be carried out, except very partially. In the science of language, genealogical cla.s.sification must rest chiefly on the formal or grammatical elements, which, after they have been affected by phonetic change, can be kept up only by a continuous tradition. We know that French, Italian, Spanish, and Portuguese must be derived from a common source, because they share grammatical forms in common, which none of these dialects could have supplied from their own resources, and which have no meaning, or, so to say, no life, in any one of them. The termination of the imperfect _ba_ in Spanish, _va_ in Italian, by which _canto_, I sing, is changed into _cantaba_ and _cantava_, has no separate existence, and no independent meaning in either of these modern dialects.

It could not have been formed with the materials supplied by Spanish and Italian. It must have been handed down from an earlier generation in which this _ba_ had a meaning. We trace it back to Latin _bam_, in _cantabam_, and here it can be proved that _bam_ was originally an independent auxiliary verb, the same which exists in Sanskrit _bhavami_, and in the Anglo-Saxon _beom_, I am. Genealogical cla.s.sification, therefore, applies properly only to decaying languages, to languages in which grammatical growth has been arrested, through the influence of literary cultivation; in which little new is added, everything old is retained as long as possible, and where what we call growth or history is nothing but the progress of phonetic corruption. But before languages decay, they have pa.s.sed through a period of growth; and it seems to have been completely overlooked, that dialects which diverged during that early period, would naturally resist every attempt at genealogical cla.s.sification. If you remember the manner in which, for instance, the plural was formed in Chinese and other languages examined by us in a former Lecture, you will see that where each dialect may choose its own term expressive of plurality, such as _heap_, _cla.s.s_, _kind_, _flock_, _cloud_, &c., it would be unreasonable to expect similarity in grammatical terminations, after these terms have been ground down by phonetic corruption to mere exponents of plurality. But, on the other hand, it would by no means follow that therefore these languages had no common origin. Languages may have a common origin, and yet the words which they originally employed for marking case, number, person, tense, and mood, having been totally different, the grammatical terminations to which these words would gradually dwindle down could not possibly yield any results if submitted to the a.n.a.lysis of comparative grammar. A genealogical cla.s.sification of such languages is, therefore, from the nature of the case, simply impossible, at least, if such cla.s.sification is chiefly to be based on grammatical or formal evidence.

It might be supposed, however, that such languages, though differing in their grammatical articulation, would yet evince their common origin by the ident.i.ty of their radicals or roots. No doubt, they will in many instances. They will probably have retained their numerals in common, some of their p.r.o.nouns, and some of the commonest words of every-day life. But even here we must not expect too much, nor be surprised if we find even less than we expected. You remember how the names for father varied in the numerous Friesian dialects. Instead of _frater_, the Latin word for brother, you find _hermano_ in Spanish. Instead of _ignis_, the Latin word for fire, you have in French _feu_, in Italian, _fuoco_. n.o.body would doubt the common origin of German and English; yet the English numeral "the first," though preserved in _Furst_, _prnceps_, prince, is quite different from the German "Der Erste;" "the second" is quite different from "Der Zweite;" and there is no connection between the possessive p.r.o.noun _its_, and the German _sein_. This dialectical freedom works on a much larger scale in ancient and illiterate languages; and those who have most carefully watched the natural growth of dialects will be the least surprised that dialects which had the same origin should differ, not only in their grammatical framework, but likewise in many of those test-words which are very properly used for discovering the relations.h.i.+p of literary languages. How it is possible to say anything about the relations.h.i.+p of such dialects we shall see hereafter. For the present, it is sufficient if I have made it clear why the principle of genealogical cla.s.sification is not of necessity applicable to all languages; and secondly, why languages, though they cannot be cla.s.sified genealogically, need not therefore be supposed to have been different from the beginning. The a.s.sertion so frequently repeated that the impossibility of cla.s.sing all languages genealogically proves the impossibility of a common origin of language, is nothing but a kind of scientific dogmatism, which, more than anything else, has impeded the free progress of independent research.

But let us see now how far the genealogical cla.s.sification of languages has advanced, how many families of human speech have been satisfactorily established. Let us remember what suggested to us the necessity of a genealogical cla.s.sification. We wished to know the original intention of certain words and grammatical forms in English, and we saw that before we could attempt to fathom the origin of such words as "I love," and "I loved," we should have to trace them back to their most primitive state.

We likewise found, by a reference to the history of the Romance dialects, that words existing in one dialect had frequently been preserved in a more primitive form in another, and that, therefore, it was of the highest importance to bring ancient languages into the same genealogical connection by which French, Italian, Spanish, and Portuguese are held together as the members of one family.

Beginning, therefore, with the living language of England, we traced it, without difficulty, to Anglo-Saxon. This carries us back to the seventh century after Christ, for it is to that date that Kemble and Thorpe refer the ancient English epic, the Beowulf. Beyond this we cannot go on English soil. But we know that the Saxons, the Angles, and Jutes came from the continent, and there their descendants, along the northern coast of Germany, still speak _Low-German_, or Nieder-Deutsch, which in the harbors of Antwerp, Bremen, and Hamburg, has been mistaken by many an English sailor for a corrupt English dialect. The Low-German comprehends many dialects in the north or the lowlands of Germany; but in Germany proper they are hardly ever used for literary purposes. The Friesian dialects are Low-German, so are the Dutch and Flemish. The Friesian had a literature of its own as early at least as the twelfth century, if not earlier.(167) The Dutch, which is still a national and literary language, though confined to a small area, can be traced back to literary doc.u.ments of the sixteenth century. The Flemish, too, was at that time the language of the court of Flanders and Brabant, but has since been considerably encroached upon, though not yet extinguished, by the official languages of the kingdoms of Holland and Belgium. The oldest literary doc.u.ment of Low-German on the Continent is the Christian epic, the _Heljand_ (Heljand = Heiland, the Healer or Saviour), which is preserved to us in two MSS. of the ninth century, and was written at that time for the benefit of the newly converted Saxons. We have traces of a certain amount of literature in Saxon or Low-German from that time onward through the Middle Ages up to the seventeenth century. But little only of that literature has been preserved; and, after the translation of the Bible by Luther into High-German, the fate of Low-German literature was sealed.

The literary language of Germany is, and has been ever since the days of Charlemagne, the _High-German_. It is spoken in various dialects all over Germany.(168) Its history may be traced through three periods. The present, or New High-German period dates from Luther; the Middle High-German period extends from Luther backwards to the twelfth century; the Old High-German period extends from thence to the seventh century.

Thus we see that we can follow the High-German, as well as the Low-German branch of Teutonic speech, back to about the seventh century after Christ.

We must not suppose that before that time there was _one_ common Teutonic language spoken by all German tribes, and that it afterwards diverged into two streams,-the High and Low. There never was a common, uniform, Teutonic language; nor is there any evidence to show that there existed at any time a uniform High-German or Low-German language, from which all High-German and Low-German dialects are respectively derived. We cannot derive Anglo-Saxon, Friesian, Flemish, Dutch, and Platt-Deutsch from the ancient Low-German, which is preserved in the continental Saxon of the ninth century. All we can say is this, that these various Low-German dialects in England, Holland, Friesia, and Lower Germany, pa.s.sed at different times through the same stages, or, so to say, the same lat.i.tudes of grammatical growth. We may add that, with every century that we go back, the convergence of these dialects becomes more and more decided; but there is no evidence to justify us in admitting the historical reality of _one_ primitive and uniform Low-German language from which they were all derived. This is a mere creation of grammarians who cannot understand a multiplicity of dialects without a common type. They would likewise demand the admission of a primitive High-German language, as the source, not only of the literary Old, Middle, and Modern High-German, but likewise of all the local dialects of Austria, Bavaria, Swabia, and Franconia. And they would wish us to believe that, previous to the separation into High and Low German, there existed one complete Teutonic language, as yet neither High nor Low, but containing the germs of both. Such a system may be convenient for the purposes of grammatical a.n.a.lysis, but it becomes mischievous as soon as these grammatical abstractions are invested with an historical reality. As there were families, clans, confederacies, and tribes, before there was a nation; so there were dialects before there was a language. The grammarian who postulates an historical reality for the one primitive type of Teutonic speech, is no better than the historian who believes in a _Francus_, the grandson of Hector, and the supposed ancestor of all the Franks, or in a _Brutus_, the mythical father of all the Britons. When the German races descended, one after the other, from the Danube and from the Baltic, to take possession of Italy and the Roman provinces,-when the Goths, the Lombards, the Vandals, the Franks, the Burgundians, each under their own kings, and with their own laws and customs, settled in Italy, Gaul, and Spain, to act their several parts in the last scene of the Roman tragedy,-we have no reason to suppose that they all spoke one and the same dialect. If we possessed any literary doc.u.ments of those ancient German races, we should find them all dialects again, some with the peculiarities of High, others with those of Low, German. Nor is this mere conjecture: for it so happens that, by some fortunate accident, the dialect of one at least of those ancient German races has been preserved to us in the Gothic translation of the Bible by Bishop Ulfilas.

I must say a few words on this remarkable man. The accounts of ecclesiastical historians with regard to the date and the princ.i.p.al events in the life of Ulfilas are very contradictory. This is partly owing to the fact that Ulfilas was an Arian bishop, and that the accounts which we possess of him come from two opposite sides, from Arian and Athanasian writers. Although in forming an estimate of his character it would be necessary to sift this contradictory evidence, it is but fair to suppose that, when dates and simple facts in the life of the Bishop have to be settled, his own friends had better means of information than the orthodox historians. It is, therefore, from the writings of his own co-religionists that the chronology and the historical outline of the Bishop's life should be determined.

The princ.i.p.al writers to be consulted are Philostorgius, as preserved by Photius, and Auxentius, as preserved by Maximinus in a MS. lately discovered by Professor Waitz(169) in the Library at Paris. (Supplement.

Latin. No. 594.) This MS. contains some writings of Hilarius, the two first books of Ambrosius De fide, and the acts of the Council of Aquileja (381). On the margin of this MS. Maximinus repeated the beginning of the acts of the Council of Aquileja, adding remarks of his own in order to show how unfairly Palladius had been treated in that council by Ambrose.

He jotted down his own views on the Arian controversy, and on fol. 282, seq., he copied an account of Ulfilas written by Auxentius, the bishop of Dorostorum (Silistria on the Danube), a pupil of Ulfilas. This is followed again by some dissertations of Maximinus, and on foll. 314-327, a treatise addressed to Ambrose by a Semi-arian, a follower of Eusebius, possibly by Prudentius himself, was copied and slightly abbreviated for his own purposes by Maximinus.

It is from Auxentius, as copied by Maximinus, that we learn that Ulfilas died at Constantinople, where he had been invited by the emperor to a disputation. This could not have been later than the year 381, because, according to the same Auxentius, Ulfilas had been bishop for forty years, and, according to Philostorgius, he had been consecrated by Eusebius. Now Eusebius of Nicomedia died 341, and as Philostorgius says that Ulfilas was consecrated by "Eusebius and the bishops who were with him," the consecration has been referred with great plausibility to the beginning of the year 341, when Eusebius presided at the Synod of Antioch. As Ulfilas was thirty years old at the time of his consecration, he must have been born in 311, and as he was seventy years of age when he died at Constantinople, his death must have taken place in 381.

Professor Waitz fixed the death of Ulfilas in 388, because it is stated by Auxentius that other Arian bishops had come with Ulfilas on his last journey to Constantinople, and had actually obtained the promise of a new council from the emperors, but that the heretical party, _i.e._, the Athanasians, succeeded in getting a law published, prohibiting all disputation on the faith, whether in public or private. Maximinus, to whom we owe this notice, has added two laws from the Codex Theodosia.n.u.s, which he supposed to have reference to this controversy, dated respectively 388 and 386. This shows that Maximinus himself was doubtful as to the exact date. Neither of these laws, however, is applicable to the case, as has been fully shown by Dr. Bessell. They are quotations from the Codex Theodosia.n.u.s made by Maximinus at his own risk, and made in error. If the death of Ulfilas were fixed in 388, the important notice of Philostorgius, that Ulfilas was consecrated by Eusebius, would have to be surrendered, and we should have to suppose that as late as 388 Theodosius had been in treaty with the Arians, whereas after the year 383, when the last attempt at a reconciliation bad been made by Theodosius, and had failed, no mercy was any longer shown to the party of Ulfilas and his friends.

If, on the contrary, Ulfilas died at Constantinople in 381, he might well have been called there by the Emperor Theodosius, not to a council, but to a disputation (ad disputationem), as Dr. Bessell ingeniously maintains, against the Psathyropolistae,(170) a new sect of Arians at Constantinople.

About the same time, in 380, Sozomen(171) refers to efforts made by the Arians to gain influence with Theodosius. He mentions, like Auxentius, that these efforts were defeated, and a law published to forbid disputations on the nature of G.o.d. This law exists in the Codex Theodosia.n.u.s, and is dated January 10, 381. But what is most important is, that this law actually revokes a rescript that had been obtained fraudulently by the Arian heretics, thus confirming the statement of Auxentius that the emperor had held out to him and his party a promise of a new council.

We now return to Ulfilas. He was born in 311. His parents, as Philostorgius tells us, were of Cappadocian origin, and had been carried away by the Goths as captives from a place called Sadagolthina, near the town of Parna.s.sus. It was under Valerian and Gallienus (about 267) that the Goths made this raid from Europe to Asia, Galatia, and Cappadocia, and the Christian captives whom they carried back to the Danube were the first to spread the light of the Gospel among the Goths. Philostorgius was himself a Cappadocian, and there is no reason to doubt this statement of his on the parentage of Ulfilas. Ulfilas was born among the Goths; Gothic was his native language, though he was able in after-life to speak and write both in Latin and Greek. Philostorgius, after speaking of the death of Crispus (326), and before proceeding to the last years of Constantine, says, that "about that time" Ulfilas led his Goths from beyond the Danube into the Roman empire. They had to leave their country, being persecuted on account of their Christianity. Ulfilas was the leader of the faithful flock, and came to Constantine, (not Constantius,) as amba.s.sador. This must have been before 337, the year of Constantine's death. It may have been in 328, when Constantine had gained a victory over the Goths; and though Ulfilas was then only seventeen years of age, this would be no reason for rejecting the testimony of Philostorgius, who says that Constantine treated Ulfilas with great respect, and called him the Moses of his time. Having led his faithful flock across the Danube into Msia, he might well have been compared by the emperor to Moses leading the Israelites from Egypt through the Red Sea. It is true that Auxentius inst.i.tutes the same comparison between Ulfilas and Moses, after stating that Ulfilas had been received with great honors by Constantius. But this refers to what took place after Ulfilas had been for seven years bishop among the Goths, in 348, and does not invalidate the statement of Philostorgius as to the earlier intercourse between Ulfilas and Constantine. Sozomen (H. E. vi. 3, 7) clearly distinguishes between the first crossing of the Danube by the Goths, with Ulfilas as their amba.s.sador, and the later attacks of Athanarich on Fridigern or Fritiger, which led to the settlement of the Goths in the Roman empire. We must suppose that after having crossed the Danube, Ulfilas remained for some time with his Goths, or at Constantinople. Auxentius says that he officiated as Lector, and it was only when he had reached the requisite age of thirty, that he was made bishop by Eusebius in 341. He pa.s.sed the first seven years of his episcopate among the Goths, and the remaining thirty-three of his life "in solo Romaniae," where he had migrated together with Fritiger and the Thervingi. There is some confusion as to the exact date of the Gothic Exodus, but it is not at all unlikely that Ulfilas acted as their leader on more than one occasion.

There is little more to be learnt about Ulfilas from other sources. What is said by ecclesiastical historians about the motives of his adopting the doctrines of Arius, and his changing from one side to the other, deserves no credit. Ulfilas, according to his own confession, was always an Arian (semper sic credidi). Socrates says that Ulfilas was present at the Synod of Constantinople in 360, which may be true, though neither Auxentius nor Philostorgius mentions it. The author of the Acts of Nicetas speaks of Ulfilas as present at the Council of Nicaea, in company with Theophilus.

Theophilus, it is true, signed his name as a Gothic bishop at that council, but there is nothing to confirm the statement that Ulfilas, then fourteen years of age, was with Theophilus.

Ulfilas translated the whole Bible, except the Books of Kings. For the Old Testament he used the Septuagint; for the New, the Greek text; but not exactly in that form in which we have it. Unfortunately, the greater part of his work has been lost, and we have only considerable portions of the Gospels, all the genuine Epistles of St. Paul, though again not complete; fragments of a Psalm, of Ezra, and Nehemiah.(172)

Though Ulfilas belonged to the western Goths, his translation was used by all Gothic tribes, when they advanced into Spain and Italy. The Gothic language died out in the ninth century, and after the extinction of the great Gothic empires, the translation of Ulfilas was lost and forgotten.

But a MS. of the fifth century had been preserved in the Abbey of Werden, and towards the end of the sixteenth century, a man of the name of Arnold Mercator, who was in the service of William IV., the Landgrave of Hessia, drew attention to this old parchment containing large fragments of the translation of Ulfilas. The MS., known as the Codex Argenteus, was afterwards transferred to Prague, and when Prague was taken in 1648 by Count Konigsmark, he carried this Codex to Upsala in Sweden, where it is still preserved as one of the greatest treasures. The parchment is purple, the letters in silver, and the MS. bound in solid silver.

In 1818, Cardinal Mai and Count Castiglione discovered some more fragments in the Monastery of Bobbio, where they had probably been preserved ever since the Gothic empire of Theodoric the Great in Italy had been destroyed.

Ulfilas must have been a man of extraordinary power to conceive, for the first time, the idea of translating the Bible into the vulgar language of his people. At his time, there existed in Europe but two languages which a Christian bishop would have thought himself justified in employing, Greek and Latin. All other languages were still considered as barbarous. It required a prophetic sight, and a faith in the destinies of these half-savage tribes, and a conviction also of the utter effeteness of the Roman and Byzantine empires, before a bishop could have brought himself to translate the Bible into the vulgar dialect of his barbarous countrymen.

Soon after the death of Ulfilas, the number of Christian Goths at Constantinople had so much increased as to induce Chrysostom, the bishop of Constantinople (397-405), to establish a church in the capital, where the service was to be read in Gothic.(173)

The language of Ulfilas, the Gothic, belongs, through its phonetic structure, to the Low-German cla.s.s, but in its grammar it is, _with few exceptions_, far more primitive than the Anglo-Saxon of the Beowulf, or the Old High-German of Charlemagne. These few exceptions, however, are very important, for they show that it would be grammatically, and therefore historically, impossible to derive either Anglo-Saxon or High-German, or both,(174) from Gothic. It would be impossible, for instance, to treat the first person plural of the indicative present, the Old High-German _nerjames_, as a corruption of the Gothic _nasjam_; for we know, from the Sanskrit _masi_, the Greek _mes_, the Latin _mus_, that this was the original termination of the first person plural.

Gothic is but one of the numerous dialects of the German race; some of which became the feeders of the literary languages of the British Isles, of Holland, Friesia, and of Low and High Germany, while others became extinct, and others rolled on from century to century unheeded, and without ever producing any literature at all. It is because Gothic is the only one of these parallel dialects that can be traced back to the fourth century, whereas the others disappear from our sight in the seventh, that it has been mistaken by some for the original source of all Teutonic speech. The same arguments, however, which we used against Raynouard, to show that Provencal could not be considered as the parent of the Six Romance dialects, would tell with equal force against the pretensions of Gothic to be considered as more than the eldest sister of the Teutonic branch of speech.

There is, in fact, a third stream of Teutonic speech, which a.s.serts its independence as much as High-German and Low-German, and which it would be impossible to place in any but a co-ordinate position with regard to Gothic, Low and High German. This is the _Scandinavian_ branch. It consists at present of three literary dialects, those of Sweden, Denmark, and Iceland, and of various local dialects, particularly in secluded valleys and fiords of Norway,(175) where, however, the literary language is Danish.

It is commonly supposed(176) that, as late as the eleventh century, identically the same language was spoken in Sweden, Norway, and Denmark, and that this language was preserved almost intact in Iceland, while in Sweden and Denmark it grew into two new national dialects. Nor is there any doubt that the Icelandic skald recited his poems in Iceland, Norway, Sweden, Denmark, nay, even among his countrymen in England and Gardariki, without fear of not being understood, till, as it is said, William introduced Welsh, _i.e._ French, into England, and Slavonic tongues grew up in the east.(177) But though one and the same language (then called Danish or Norraenish) was understood, I doubt whether one and the same language was spoken by all Northmen, and whether the first germs of Swedish and Danish did not exist long before the eleventh century, in the dialects of the numerous clans and tribes of the Scandinavian race. That race is clearly divided into two branches, called by Swedish scholars the East and West Scandinavian. The former would be represented by the old language of Norway and Iceland, the latter by Swedish and Danish. This division of the Scandinavian race had taken place before the Northmen settled in Sweden and Norway. The western division migrated westward from Russia, and crossed over from the continent to the Aland Islands, and from thence to the southern coast of the peninsula. The eastern division travelled along the Bothnian Gulf, pa.s.sing the country occupied by the Finns and Lapps, and settled in the northern highlands, spreading toward the south and west.

The earliest fragments of Scandinavian speech are preserved in the two _Eddas_, the elder or poetical Edda, containing old mythic poems, the younger or Snorri's Edda giving an account of the ancient mythology in prose. Both Eddas were composed, not in Norway, but in Iceland, an island about as large as Ireland, and which became first known through some Irish monks who settled there in the eighth century.(178) In the ninth century voyages of discovery were made to Iceland by Naddodd, Gardar, and Flokki, 860-870, and soon after the distant island, distant about 750 English miles from Norway, became a kind of America to the Puritans and Republicans of the Scandinavian peninsula. Harald Haarf.a.gr (850-933) had conquered most of the Norwegian kings, and his despotic sway tended to reduce the northern freemen to a state of va.s.salage. Those who could not resist, and could not bring themselves to yield to the sceptre of Harald, left their country and migrated to France, to England, and to Iceland (874). They were mostly n.o.bles and freemen, and they soon established in Iceland an aristocratic republic, such as they had had in Norway before the days of Harald. This northern republic flourished; it adopted Christianity in the year 1000. Schools were founded, two bishoprics were established, and cla.s.sical literature was studied with the same zeal with which their own national poems and laws had been collected and interpreted by native scholars and historians. The Icelanders were famous travellers, and the names of Icelandic students are found not only in the chief cities of Europe, but in the holy places of the East. At the beginning of the twelfth century Iceland counted 50,000 inhabitants. Their intellectual and literary activity lasted to the beginning of the thirteenth century, when the island was conquered by Hakon VI., king of Norway. In 1380, Norway, together with Iceland, was united with Denmark; and when, in 1814, Norway was ceded to Sweden, Iceland remained, as it is still, under Danish sway.

The old poetry which flourished in Norway in the eighth century, and which was cultivated by the skalds in the ninth, would have been lost in Norway itself had it not been for the jealous care with which it was preserved by the emigrants of Iceland. The most important branch of their traditional poetry were short songs (hliod or Quida), relating the deeds of their G.o.ds and heroes. It is impossible to determine their age, but they existed at least previous to the migration of the Northmen to Iceland, and probably as early as the seventh century, the same century which yields the oldest remnants of Anglo-Saxon, Low-German, and High-German. They were collected in the middle of the twelfth century by _Saemund Sigfusson_ (died 1133).

In 1643 a similar collection was discovered in MSS. of the thirteenth century, and published under the t.i.tle of _Edda_, or Great-Grandmother.

This collection is called the old or poetic Edda, in order to distinguish it from a later work ascribed to Snorri Sturluson (died 1241). This, the younger or prose Edda, consists of three parts: the mocking of Gylfi, the speeches of Bragi, and the Skalda, or _Ars poetica_. Snorri Sturluson has been called the Herodotus of Iceland; and his chief work is the "Heimskringla," the world-ring, which contains the northern history from the mythic times to the time of King Magnus Erlingsson (died 1177). It was probably in preparing his history that, like Ca.s.siodorus, Saxo Grammaticus, Paulus Diaconus, and other historians of the same cla.s.s, Snorri collected the old songs of the people; for his "Edda," and particularly his "Skalda," are full of ancient poetic fragments.

The "Skalda," and the rules which it contains, represent the state of poetry in the thirteenth century; and nothing can be more artificial, nothing more different from the genuine poetry of the old "Edda" than this _Ars poetica_ of Snorri Sturluson. One of the chief features of this artificial or skaldic poetry was this, that nothing should be called by its proper name. A s.h.i.+p was not to be called a s.h.i.+p, but the beast of the sea; blood, not blood, but the dew of pain, or the water of the sword. A warrior was not spoken of as a warrior, but as an armed tree, the tree of battle. A sword was the flame of wounds. In this poetical language, which every skald was bound to speak, there were no less than 115 names for Odin; an island could be called by 120 synonymous t.i.tles. The specimens of ancient poetry which Snorri quotes are taken from the skalds, whose names are well known in history, and who lived from the tenth to the thirteenth century. But he never quotes from any song contained in the old "Edda,"(179) whether it be that those songs were considered by himself as belonging to a different and much more ancient period of literature, or that they could not be used in ill.u.s.tration of the scholastic rules of skaldic poets, these very rules being put to shame by the simple style of the national poetry, which expressed what it had to express without effort and circ.u.mlocution.

We have thus traced the modern Teutonic dialects back to four princ.i.p.al channels,-the _High-German_, _Low-German_, _Gothic_, and _Scandinavian_; and we have seen that these four, together with several minor dialects, must be placed in a co-ordinate position from the beginning, as so many varieties of Teutonic speech. This Teutonic speech may, for convenience'

sake, be spoken of as one,-as one branch of that great family of language to which, as we shall see, it belongs; but it should always be borne in mind that this primitive and uniform language never had any real historical existence, and that, like all other languages, that of the Germans began with dialects which gradually formed themselves into several distinct national deposits.

We must now advance more rapidly, and, instead of the minuteness of an Ordnance-map, we must be satisfied with the broad outlines of Wyld's Great Globe in our survey of the languages which, together with the Teutonic, form the Indo-European or Aryan family of speech.

And first the Romance, or modern Latin languages. Leaving mere local dialects out of sight, we have at present six literary modifications of Latin, or more correctly, of ancient Italian,-the languages of Portugal, of Spain, of France, of Italy, of Wallachia,(180) and of the Grisons of Switzerland, called the Roumansch or Romanese.(181) The Provencal, which, in the poetry of the Troubadours, attained at a very early time to a high literary excellence, has now sunk down to a mere _patois_. The earliest Provencal poem, the Song of Boethius, is generally referred to the tenth century: Le Buf referred it to the eleventh. But in the lately discovered Song of Eulalia, we have now a specimen of the Langue d'Oil, or the ancient Northern French, anterior in date to the earliest poetic specimen of the Langue d'Oc, or the ancient Provencal. Nothing can be a better preparation for the study of the comparative grammar of the ancient Aryan languages than a careful perusal of the "Comparative Grammar of the Six Romance Languages" by Professor Diez.

Though in a general way we trace these six Romance languages back to Latin, yet it has been pointed out before that the cla.s.sical Latin would fail to supply a complete explanation of their origin. Many of the ingredients of the Neo-Latin dialects must be sought for in the ancient dialects of Italy and her provinces. More than one dialect of Latin was spoken there before the rise of Rome, and some important fragments have been preserved to us, in inscriptions, of the Umbrian spoken in the north, and of the Oscan spoken to the south of Rome. The Oscan language, spoken by the Samnites, now rendered intelligible by the labors of Mommsen, had produced a literature before the time of Livius Andronicus; and the tables of Iguvio, so elaborately treated by Aufrecht and Kirchhoff, bear witness to a priestly literature among the Umbrians at a very early period. Oscan was still spoken under the Roman emperors, and so were minor local dialects in the south and the north. As soon as the literary language of Rome became cla.s.sical and unchangeable, the first start was made in the future career of those dialects which, even at the time of Dante, are still called _vulgar_ or _popular_.(182) A great deal, no doubt, of the corruption of these modern dialects is due to the fact that, in the form in which we know them after the eighth century, they are really Neo-Latin dialects as adopted by the Teutonic barbarians; full, not only of Teutonic words, but of Teutonic idioms, phrases, and constructions. French is provincial Latin as spoken by the Franks, a Teutonic race; and, to a smaller extent, the same _barbarizing_ has affected all other Roman dialects. But from the very beginning, the stock with which the Neo-Latin dialects started was not the cla.s.sical Latin, but the vulgar, local, provincial dialects of the middle, the lower, and the lowest cla.s.ses of the Roman Empire. Many of the words which give to French and Italian their cla.s.sical appearance, are really of much later date, and were imported into them by mediaeval scholars, lawyers, and divines; thus escaping the rough treatment to which the original vulgar dialects were subjected by the Teutonic conquerors.

The next branch of the Indo-European family of speech is the _h.e.l.lenic_.

Its history is well known from the time of Homer to the present day. The only remark which the comparative philologist has to make is that the idea of making Greek the parent of Latin, is more preposterous than deriving English from German; the fact being that there are many forms in Latin more primitive than their corresponding forms in Greek. The idea of Pelasgians as the common ancestors of Greeks and Romans is another of those grammatical mythes, but hardly requires at present any serious refutation.

The fourth branch of our family is the _Celtic_. The Celts seem to have been the first of the Aryans to arrive in Europe; but the pressure of subsequent migrations, particularly of Teutonic tribes, has driven them towards the westernmost parts, and latterly from Ireland across the Atlantic. At present the only remaining dialects are the Kymric and Gadhelic. The _Kymric_ comprises the _Welsh_; the _Cornish_, lately extinct; and the _Armorican_, of Brittany. The _Gadhelic_ comprises the _Irish_; the _Galic_ of the west coast of Scotland; and the dialect of the _Isle of Man_. Although these Celtic dialects are still spoken, the Celts themselves can no longer be considered an independent nation, like the Germans or Slaves. In former times, however, they not only enjoyed political autonomy, but a.s.serted it successfully against Germans and Romans. Gaul, Belgium, and Britain were Celtic dominions, and the north of Italy was chiefly inhabited by them. In the time of Herodotus we find Celts in Spain; and Switzerland, the Tyrol, and the country south of the Danube have once been the seats of Celtic tribes. But after repeated inroads into the regions of civilization, familiarizing Latin and Greek writers with the names of their kings, they disappear from the east of Europe. Brennus is supposed to mean king, the Welsh _brennin_. A Brennus conquered Rome (390), another Brennus threatened Delphi (280). And about the same time a Celtic colony settled in Asia, and founded Galatia, where the language spoken at the time of St. Jerome was still that of the Gauls.

Celtic words may be found in German, Slavonic, and even in Latin, but only as foreign terms, and their amount is much smaller than commonly supposed.

A far larger number of Latin and German words have since found their way into the modern Celtic dialects, and these have frequently been mistaken by Celtic enthusiasts for original words, from which German and Latin might, in their turn, be derived.

The fifth branch, which is commonly called _Slavonic_, I prefer to designate by the name of _Windic_, _Winidae_ being one of the most ancient and comprehensive names by which these tribes were known to the early historians of Europe. We have to divide these tribes into two divisions, the _Lettic_ and the _Slavonic_, and we shall have to subdivide the Slavonic again into a _South-East Slavonic_ and a _West Slavonic_ branch.

The _Lettic_ division consists of languages hardly known to the student of literature, but of great importance to the student of language. _Lettish_ is the language now spoken in Kurland and Livonia. _Lithuanian_ is the name given to a language still spoken by about 200,000 people in Eastern Prussia, and by more than a million of people in the coterminous parts of Russia. The earliest literary doc.u.ment of Lithuanian is a small catechism of 1547.(183) In this, and even in the language as now spoken by the Lithuanian peasant, there are some grammatical forms more primitive, and more like Sanskrit, than the corresponding forms in Greek and Latin.

The _Old Prussian_, which is nearly related to Lithuanian, became extinct in the seventeenth century, and the entire literature which it has left behind consists in an old catechism.