CHAPTER II.

The Aryan Languages.

§ 31. From the one primitive fundamental form we pass 173 on to manifold later forms, which have arisen from various changes of the former, according to the doctrine which prevails in modern philology and which we have just seen to be much more probable than the contrary theory. Between that one form and these manifold forms of Aryan speech what are the relations existing as far as concerns the method and the order of their development? Besides the bond of kinship which joins together all the families of the IndoEuropean languages as being all descended from he prehistoric Aryan, are there bonds which unite more closely family to family? And, if there are, in what way ought we to conceive of the division and the subdivisions of the mother-language of all the European languages? This is the last problem to which we shall invite the attention of our readers.

Its solution, one most important for philology and history, has been attempted, even before the last decade, by learned investigators, among whom it will suffice to mention Schleicher and Lottner. The former by a series of comparative researches, for the most part in a strict sense grammatical (that is to say, phonological and morphological) the results of which he published in various works,' was led to divide the Aryan stock into two great sections, the first of which we will call Indo-Irano-Graeco-Italo-Keltic, the 173

1 Kurzer abriss der geschichte stamme (ibid., i. 437-48). - Die der slavischen sprache (Beiträge, etc., i. 1-27). Die stellung des celtischen im indogermanischen sprach

deutsche sprache, etc., pp. 80-2 (edition of 1869).—Compendium, etc. pp.5-9.

second Lithu-Slavo-Teutonic. The first is subdivided into Indo-Iranic (which Schleicher calls 'Aryan,' understanding the word in a narrower sense than that in which we use it) and Graeco-Italo-Keltic: the Indo-Iranic into the Indian and the Iranic family; the Graeco-Italo-Keltic into the Greek (with which he connects Albanian) and the ItaloKeltic family; this last into the Italic and the Keltic family. Similarly, the second section is subdivided into Slavo-Lithuanian and Teutonic; the Slavo-Lithuanian into the Slavonic and the Lithuanian family.-The second of the two philologists mentioned, Lottner,' also sees in the class of the Aryan languages, to begin with, two sections quite distinct from each other, the Asiatic and the European. From the first are derived the Indian and the Iranic family. From the second was divided off primarily, as it seems, Greek (or Helleno-Phrygian): the remainder might have been split into two portions, a south-western and a northern; the first might have given rise to Keltic and Italic; the second to German and to Slavo-Lithuanian, which was afterwards subdivided into two families of language.3

1 Über die stellung der Italer innerhalb des indo-europäischen stammes (Zeitschr. f. vgl, sprachforsch,vii. 18-49, 161-93).-Celtischitalisch (Beiträge, etc., ii. 309-21.)

2 The characteristics which, according to this author, distinguish the European Aryan from the Asiatic Aryan are the following: 1st, the change of r to l, which has taken place in a large number of material elements and formal elements; 2nd, the loss of aspiration, a phenomenon common to the Aryan languages of Europe in certain words; 3rd, the well-marked meaning which in these languages has been assumed by several prepositions which do not show themselves endowed with a

like force in Sanscrit ; 4th, the prehistoric civilisation of the European Aryans, as it appears to us from a series of words found in all their families of languages.-The division of Aryan into Asiatic and European is favoured also by Curtius's investigations into the sounds which succeeded the primitive a (Über die spaltung des a-lautes, etc., in the Sitzungberichte der K. sächs. Gessellsch. d. wissensch., 1864, p. 9, sqq.).

3 By grammatical and lexical ob. servations Lottner endeavours to prove that the Italic family shows itself to be akin not so much to the Greek as to the Northern languages of the European section. He after

In the division of Aryan into Asiatic and European 174 Scherer, Fick, and other philologists agree with Lottner. Scherer in the introduction to his work Zur geschichte der deutschen sprache (p. 4) declares his opinion that this division of Aryan into Eastern and Western, as he expresses himself, is proved. Fick, after having attempted in his Vergleichendes wörterbuch der indogermanischen sprachen1 to give a complete lexical demonstration of this doctrine, defended it against the objections of J. Schmidt, of whom we shall speak directly, in the book entitled Die ehemalige spracheinheit der Indogermanen Europas, which (as well as the preceding) we have already had occasion to touch upon. The principal arguments which Fick adduced in favour of the division above mentioned of Aryan are three: 1st, the development of e from original a, a development which

wards proceeds to demonstrate the same position with respect to Keltic. He founds his arguments especially on the structure of the verb in the languages alluded to. The history of European Aryan after the separation from it of Greek is thus described by Lottner: "the old tonic law gradually loses all force; in conjugation the augment disappears, and is compensated for in the imperfect by means of a species of internal or final augment; . . . . the conjunction gives up its ancient field to the potential invader: the middle begins to disappear entirely;

on the other hand the tenth class becomes a source of prolific new formations."

With respect to the relations of Keltic with the other Aryan languages see also the very weighty opinion of Ebel (Celtisch, griechisch, lateinisch, in the Beiträge, etc., i. 429-37.-Die stellung des celt

ischen, ibid., ii. 137-94); according to this illustrious philologist Keltic would occupy an intermediate position between Teutonic and Italic, but shares more with the former than with the latter the characteristics belonging to the innermost part of the language.

1 See the second and third sections, in which are collected the words belonging to the Indo-Iranic linguistic unity and the European linguistic unity: see also pp. 1045-56 of the second edition.

2 On this work see the reflections of J. Schmidt (in the Jenaer literaturzeitung, 1874, pp. 201-4) and of Havet (in the Revue critique d'histoire et de littérature, 8th year, 1st semester, pp. 145-50, 239-40). See also Jolly, Noch einmal der stammbaum der indogermanischen sprachen (Zeitschrift. f. völkerpsychologie, etc., viii. 190205).

is common to all the Aryan languages of Europe and distinguishes them from the Indo-Iranic; 2nd, the change in not a few cases of Proto-Aryan r into / in the Aryan languages of Europe; 3rd, new formations of words and per175 haps also of roots common to these and foreign to the IndoIranic languages: words, from which we may clearly discern a difference between the civilisation of the primitive Aryans while yet undivided and that of the European Aryans as constituting a single nation; because this second

1 See above, p. 20, note 1.— Havet (ibid.) regards as still more considerable than it appears to Fick the number of instances where European e comes from an original a. This argument of Fick's has lately been subjected to a severe examination by J. Schmidt (Was beweist das e europäischen sprachen für die annahme einer einheitlichen europäischen grundsprache? in the Zeitschr. f. vgl. sprachforsch., xxiii. 333-75). Availing himself of the results of Hübschmann's latest studies on Armenian (which will be noticed in the following paragraplı) Schmidt begins by noticing that Armenian exhibits e and ei in place of original a, ai in the same words in which the Aryan languages of Europe also show e, ei. Further the e in question might have been developed not in a primitive and fundamental European language, but in fact on one soil or several spots of the European region, and hence have been broadly propagated by degrees. And that this is the most probable hypothesis appears from the fact, admitted also by Fick, that the e became increasingly frequent in each of the several Aryan anguages of Europe in the period s which followed the division of Proto

European. Then Schmidt proceeds to a careful examination of a large number of European words which contain an e = orig. a in radical or suffix syllables and comes to the two following conclusions: 1st, side by side with the forms in e uf not a few European words have been sporadically preserved also in European languages the old forms

in a;

2nd, eveu in those words which, in all the Aryan languages of Europe exhibit the original a changed, the degree of change was so various that the changed vowel in the Prussian of the middle of the 16th century was still more near to the original a than to the corresponding changed vowel of Attic, older by 2000 years. Hence, according to Schmidt, the European e= orig. a is by no means a substantial argument in favour of the existence of a fundamental European language, whence the individual Aryan languages of Europe might to have spring.

2 The force of this argument is completely destroyed or at all events greatly reduced by the researches of Heymann, who, as has been said on pp. 22-25, maintained the claims of l to be Proto-Aryan.

1

civilisation reveals itself to us in almost all its elements superior to the first, especially in the advanced agriculture. To these main arguments Fick subjoins certain others, among which it is well not to pass over altogether the formation of a class of presents with the suffix -ta-, which does not belong in this function to Indo-Iranic, and the European ho as opposed to the Asiatic k. From these facts our author draws the conclusion that the modern Aryan languages and the modern Aryan nations of Europe came from one language and a single nation. This statement must not however be understood in too strict a sense; the doctrine of which Fick made himself the champion does not exclude the existence of prehistoric differences; the unity of the primitive and fundamental Aryan language of Europe consisted especially in certain new linguistic tendencies which came to prevail in the combination of the Aryans established in this part of the world: the unity of the primitive and fundamental Aryan nation of Europe was formed by the unity of the language. Nor would it be of any use to object that, in so ancient an epoch, the Aryans of Europe could hardly have been united together into a single nation, as it seems that the formation of so great a unity implies a stage of civilisation superior to any that can be attributed to them at such an epoch: this objection cannot do away 176 with the linguistic arguments above quoted and comes into collision with history, which furnishes evidence that the Indo-Europeans, even in very remote ages, were not wanting in the capacity for combining themselves into great nations by the bond of a common language. From the denominations of the soil, the plants and the animals

1 See the first two paragraphs of this book, from which it is plain that the problem of the Proto-Aryan k has not yet been solved in a perfectly satisfactory manner. Whence

it is clear what importance should be assigned to this last proof alleged by Fick in defence of his European linguistic unity.