CH. II, (ii.) South Indo-Germanic tongue altogether, is the old Indian, the language of the oldest portion of the Vedas; at a later time in a simpler form and as a grammatical literary language, contrasted with the popular dialects, named Sanskrit. "We do not know Eranian in its original form; the oldest Eranian languages which have reached us are the old Bactrian or Zend (the eastern), and the old Persian, the language of the Achaemenidean cuneiform inscriptions (the western). To this family belongs also the Armenian, which we first know at a later time, and which must have separated earlier from the Eranian original language. "II. The south-western European portion consisting of European. (1) the Greek, nearest to which stands a language only known in its modern form, the Albanian: (2) the Italian; the oldest known forms of this family are the Latin,-and especially important for us is the old Latin, as it was spoken before the introduction of the educated literary language moulded by Greek influence, the Umbrian, and the Oscan: (3) the Keltic: the best preserved, but still very decomposed, language of the Keltic family is the Old Irish, reaching from the 7th century of our era. (iii.) North "III. The northern European portion, consisting of the European. Sclavonic family, with the closely allied Lithuanian (which is for us the important language among this group), and the Teutonic, which is widely sundered from both. 'The oldest forms of language in this portion are the Old Bulgarian-old Ecclesiastical Sclavonic in MS., dating up to the 11th century: the Lithuanian-first brought under our notice three hundred years ago, but clearly of much higher antiquity-and the Gothic, of the fourth century. Near to the Gothic, however, are the most ancient representatives of the German and the Norse, the Old High-German and Old Norse, to be brought forward where they present older forms than the Gothic. "It is in the Asiatic division that is contained most that is ancient in the sounds and in the fabric of language, and here again especially in the Old Indian. Then follows with reference to antiquity-that is to say, in the retaining its similarity to the original language, in having fewer strongly developed individual forms-the southern European division, in which the Greek had remained closest to the original; finally, the northern European group, which, taken as a whole, presents itself as developed with the most individuality, and in which the least remains of the original speech are to be traced. "If we combine this statement with the relationship already described of the Indo-Germanic languages among themselves, and draw from the two our conclusion as to the process of the divisions of the main body of IndoGermanic speech in the earliest times, we arrive at the following results: The Indo-Germanic original speech divided itself, first, by the unequal development in different parts of its province, into two sections: it divided off from itself the Sclavo-Teutonic, the language which afterwards divided into Teutonic and Sclavo-Lithuanian: and later, that portion of the original speech which remained, the Aryo-Graeco-Italo-Keltic, divided itself into GraecoItalo-Keltic and Aryan, of which the first-named soon divided itself into Greek and Italo-Keltic: and the latter, the Aryan, remained undivided for a considerable time. At a later period the Sclavo-Lithuanian, the Aryan (Indo-Eranian), and Italo-Keltic further divided themselves. It is possible that at some or all of the divisions more languages arose than are now manifest, as in many instances in process of time Indo-Germanic languages have probably become extinct. The more towards the East an Indo-Germanic people lives, so much more of what is ancient has their language retained. The more towards. the west they have gone, so much the less of what is old, and so many more new formations are to be found in their language. From these and other intimations we may conclude that the Sclavo-Teutonic race first began their CH. I'. Their de grees of relation ship. CH. II. journeyings towards the west: then followed the GræcoItalo-Keltic of the Aryans who remained behind, the Indians travelled south-eastward, and the Eranians spread in a south-westerly direction. The home of the original Indo-Germanic race is to be sought in the central highlands of Asia. "It is only of the Indians, who were the last to separate from the parent stem, that we can say with any certainty that they drove out an aboriginal people from their later dwelling-place, much of whose language passed into their own; of many of the other Indo-Germanic peoples such an hypothesis is highly probable." Prof. Schleicher proceeds to show the degrees of relationship of the main families of the Indo-Germanic speech by the diagram given below; in which the length of the lines indicates the probable time of separation. Connection between the Latin and Keltic as given by Schleicher. I have given above Schleicher's view of the near connection of the Keltic with the Italian, which is disputable. I may briefly give here some of the principal arguments on both sides, though their force will not be seen without some knowledge of the phonetic laws described in the fol lowing chapters. Schleicher believes in a "Graeco-ItaloKeltisch" period: in which the ancestors of those peoples divided the a sound into a, e, o, after parting from the Teutonic race, or at least the Gothic division of it. Then the Greeks parted off, and an Italo-Keltic period followed, distinguished by the loss of aspirates and retention of spirants, and notably also by the loss of the old middle voice and the formation of a quite new form peculiar to the Italians and the Kelts: compare legitur, Keltic legthar, with Xéуeral. After their final separation the Kelts lost the ablative and the reduplicated perfect, losses which distinguish Keltic from Italian. Other points of agreement between the Keltic and Italian, not found between any other two languages, are the formative suffix -tion (-sion), and perhaps -tric; the dative plural in b, fratribus, braithrib, while all the North European languages have m, e.g. Gothic brōthrum, the termination i alike for the genitive singular and nominative plural of the a-stem, and the future suffix -bo, -bis, &c., for which forms in b and ƒ appear in old Irish. Lottner and Ebel, on the other hand, connect the Keltic with the North European languages1. They argue from the agreement of diphthongs (ai, oi, au, iu, in Keltic, ai, ei, au, iu, in Teutonic, four in each language, while the Graeco-Italians certainly had six, ai, ei, oi, au, eu, ou): and from the different origin of some consonants; thus the Kymric ch is from h, a substitute for s, the Gaelic ƒ is a hardened v, the Kymric f, according to Ebel, is derived from s, except where it occurs in words certainly borrowed from the Latin; on the other hand the Latin h and ƒ come from aspirates, not from other spirants. 1 Ebel's arguments may be seen in Keltische Studien (Engl. trans., pp. 119-132). See also Schleicher's Kurzer Abriss der Gesch. der Ital. Sprachen in the Rheinisches Museum for 1859, and Ebel in Kuhn and Schleicher's Beiträge, 1. 429. 2 Keltic comprises the Kymric or Welsh, the now extinct Cornish, and the Armorican, or ancient language of Brittany: these three are nearly related, and are sometimes all included under the name Kymric. More distinct are the Erse or old Irish, the Gaelic of the Highlands, and the Manx: these are all sometimes called Gadhelic. CH. II. Counter arguments of Ebel. CH. II. Nature of the argu ments. Some agreements in inflexion between the Keltic and Lithuanian are not nearly so convincing as Schleicher's which are given above. Lastly, Ebel sees "a pervading analogy in the Sclavonian, Teutonic, and both branches of the Keltic1," evidenced, for instance, by the employment of prefixes to express completed action, instead of reduplication, as in Graeco-Italian; such prefixes are ru or ro in Keltic, ga in Gothic, the modern German ge, both of which have this force, though also some others. The arguments on both sides, it will be observed, are confined to the forms and inflexions of words: they are not drawn from the common possession (which is indubitable) of very many words by the Latin and Keltic, especially the Kymric. The reason is that it is generally impossible to distinguish between the genuinely Keltic words. and those which were only borrowed from the Latin after the Romans came into contact with the Kelts. When we find words like fin and flam, occurring only in Kymric and Armorican, there can be no doubt that we have here finis and flamma borrowed. But when we find words like "traeth," a sandy flat, occurring in Kymric, and in slightly different forms in Cornish, Armorican, Irish, and Gaelic, it seems unlikely that each of these races, which were probably separate before the Romans came into contact with them, either independently borrowed the Latin "tractus," or passed it on froin one to another. Still Ebel's list of borrowed words (in which traeth occurs) cannot often be challenged, and it is incomparably larger than that of words which are peculiar to the Latin and Keltic, and not shared by the North Europeans. We are therefore forced back upon the arguments from forms given above. Now such arguments are, generally speaking, stronger than any mere agreement of words. But in this case they lose much of their usual force from the obviously late character of a great part of Keltic grammar. Thus the personal suffixes of the Welsh verbs have hardly anything in common 1 Celtic Studies, p. 128. |