Like Jones's Discourse, Friedrich von Schlegel's Über die Sprache und Weisheit der Indier prepares for the important early works in nineteenth-century linguistics. Schlegel's aim too is to encourage general study of antiquity, not only of language; the section on language makes up only approximately a fourth of his book, which goes on to deal with other "media of satisfying our curiosity concerning . . . the early age" of mankind. Schlegel's book was important for arousing interest in Sanskrit, especially in Germany; it also makes the first mention of aims that were to be central to linguistics, notably "comparative grammar". Because its prime importance is its encouragement to others, only excerpts are given here, though the entire book is delightful to read. I have included one paragraph of citations comparing the vocabulary of Sanskrit and German; it may illustrate the advances made over Schlegel by his successors. And his lists of grammatical criteria for establishing relationships illustrate the enthusiasm of a popularizer rather than the care of a scholar. From the selections translated, students may learn to know the contributions of Schlegel's work as well as its shortcomings.
With his successors, Schlegel is interested in finding a common source for the languages which after Jones were held to be related. In interpreting the early conception of "source" or "derived from" we must be careful to avoid our own definitions, which are based largely on the work of subsequent linguists. In his excellent introduction to the centenary edition of Rask, Ausgewählte Abhandlungen XIII-LXIII, Holger Pedersen discusses sympathetically the use of these notions at the beginning of the nineteenth century for determining the relationship of languages. Schlegel indeed speaks of a family-tree, but derives the European languages from Sanskrit on the basis of its greater antiquity, not by positing intermediate stages. Accordingly, the relationship he suggests between German and Sanskrit should not be equated with our deriving German from Proto-Indo-European.
Schlegel's emphasis on grammar in determining relationships merits great credit. His demand for precise agreement of vocabulary items may be understood when we compare the fanciful etymologies of his predecessors; insistence on rigor was essential to stop further such fabrications. Yet while he asks for complete agreement in determining cognates, Schlegel permits the use of forms which differ, though he has not yet hit upon the concept of determining "rules" for such differences; his citing of an "analogy" between Latin p and Germanic f, Latin c and Germanic h, is a step on the way to the more comprehensive sets of rules given by Rask and Grimm.
Schlegel also is applauded for introducing the term "comparative grammar" into linguistics. In basing this term on comparative anatomy and incorporating the notion of family trees for languages, he drew on biology for linguistic methodology, foreshadowing Schleicher and his reliance on Darwinism. These adoptions of methodology and the attention he drew to Sanskrit are the most important contributions of his book.
Yet in it Schlegel also suggested a further means for distinguishing language interrelationships, one that was not taken over by Bopp, Grimm and their successors, and subsequently remained peripheral to the central course of nineteenth century linguistics: the use of typology. For Schlegel there was an ancient grammar, characterized by inflection, and a more recent grammar, characterized by analytic devices. Languages of the ancient type were more kunstreich (ingenious, artistic) than are those of the newer manner. Accordingly, examination of the type of a language might contribute to determining its antiquity. Yet in dealing with Chinese, for even Schlegel this means was disappointing; the problems of typology interested some linguists of the nineteenth century, notably Humboldt and Steinthal, but its uncertainties gave it more status among nonspecialists than among linguists. Even the efforts of Sapir in this century did little to encourage its application, though recent techniques may make it more useful. (See my Historical Linguistics, Chapter III.) Yet all attempts to use typology in support of genealogical classification have been completely discredited. We may wonder whether the ineffectiveness of typology as a tool for supporting genealogical classification led Schlegel's successors to disregard his interest in structure, which we find duplicated only in this century.
Apart from his book of 1808, the chief concern of Friedrich von Schlegel (1772-1829) was for literature. He began his study of Sanskrit and Indian antiquity in 1803, under Alexander Hamilton in Paris, planning a chrestomathy printed in Devanagari, but for it he lacked the necessary funds. Instead he published his book to arouse interest in Indic studies, expecting for European scholarship results comparable to those produced by the study of Greek in the fifteenth and sixteenth centuries. Yet after publication of the book, he himself abandoned his concern for Indic studies, in a shift of interest that may be reflected by his joining the Roman Catholic Church. From then to his death he directed his attention to Europe and his own literary production. His brother, August Wilhelm von Schlegel (1767-1845), who also concerned himself with the typological classification of language, came to concentrate on Indic studies, especially after he was appointed professor of literature at the University of Bonn in 1818. His work in this position is generally credited as the beginning of Indic scholarship in Germany. Apart from the contributions which increased knowledge of Indic languages made to linguistics, the importance of the brothers for linguistics is based almost entirely on Friedrich's book of 1808.
The Old Indic Sanskrito, that is the cultivated or perfected, also Gronthon, that is the written language or the book language, has a very close relationship with the Roman and Greek, as with the Germanic and Persian languages. The similarity consists not only in a great number of roots, which it shares with them, but it extends to the innermost structure and grammar. The agreement is accordingly not one of chance which might be explained by mixture, but rather an essential one which points to common descent. Comparison yields the further result that the Indic language is the older, the others however later and derived from it. For Armenian, the Slavic languages and next for Celtic, the relationship with Indic is either minute, or not to be compared with the close agreement among the languages named above which we derive from it. Yet this relationship, though minute, is not to be completely disregarded, since it manifests itself in accordance with the sequence in which these languages were named at least in some grammatical forms, in such components which cannot be reckoned among the chance features of the language but rather belong to their inner structure.
In Hebrew and related dialects, as well as in Coptic, a goodly number of Indic roots may possibly be found still. But this does not prove an original relationship since it can be the result of simple mixture. The grammar of these languages like that of Basque is basically different from that of Indic.
The large and not yet completely determinable number of the remaining north and south Asiatic and American Ianguages has absolutely no essential relationship with the Indic language family. To be sure, in the grammar of these languages, which also is quite different from that of Indic, we find a similar arrangement among several; in their roots however they are also completely different, even among one another and so totally deviant, that there is no possibility of being able to take them back to a common source.
The important results of this linguistic comparison for the oldest history of the origin of peoples and their earliest migrations will be the subject of investigation in the future. In this first book we will be content with establishing and making clear the principles themselves, simple but very comprehensive results of conscientious research....
Some examples may show most clearly that the claimed relationship does not in any sense rest on etymological elaborations, many of which were contrived before the proper source was found, but that it may be presented to impartial scholars as simple fact.
In making this demonstration we permit absolutely no rules of change or replacement of letters, but rather demand complete equivalence of the word as proof of descent. To be sure, if the intermediate steps can be proved historically, then giorno may be derived from dies; and if instead of Latin f we often find Spanish h, if Latin p very often becomes f in the Germanic form of the same word and Latin c not infrequently h, this certainly establishes an analogy, also for other not quite such apparent cases. Yet as indicated, one must be able to demonstrate the intermediate steps or the general analogy historically; nothing can be fabricated from axioms, and the agreement must be very precise and evident in order to permit even the minutest variations of form.
I cite first of all some Indic words which are characteristic of Germanic. Shrityoti - er schreitet 'strides'; vindoti - er findet 'finds'; schlißyoti - er umschließt 'surrounds'; Onto - das Ende 'end'; Monuschyo - der Mensch 'human being'; Shvosa, Svostri - die Schwester 'sister'; Rotho - das Rad 'wheel'; Bhruvo - die Brauen der Augen 'eyebrows'; Torsho - der Durst 'thirst'; Tandovon - der Tanz 'dance'; Ondoni - die Enten 'ducks'; Noko - der Nagel 'nail'; sthiro - unbeweglich, stier 'immovable'; Oshonon - das Essen 'food', etc....
Might one however not possibly reverse this whole proof and say: the relationship is striking enough and may be established in part; but what really is the reason for assuming that Indic is the older among the related languages and their common source? May it not just as well have arisen only through mixture of the others, or at any rate have received its similarity in this way?
Not to mention that much of what has already been mentioned and also many another probability speaks against that, we will now come to something that decides the situation fully and raises it to certainty. In general the hypothesis that attempts to derive whatever Greek elements are found in India from the Seleucids in Bactria is not much happier than one which might try to explain the Egyptian pyramids from natural crystallization.
The decisive point however which will clarify everything here is the inner structure of the languages or comparative grammar, which will give us quite new information about the genealogy of 1anguages in a similar way as comparative anatomy has illuminated the higher natural history.
Of the related languages we will first select Persian, whose grammar, which has even taken over personal suffixes from Arabic through the long and old intercourse between both peoples, agrees with that of Indic and the others far less than even that of German today, not to speak of Greek and Roman. But if one assembles all similarities, they are certainly weighty.... [to p. 32]
In Germanic grammar there are many other agreements with the Greek and Indic besides those which it shares with Persian. In Germanic as throughout in Indic, n is characteristic of the accusative, s of the genitive. The final syllable -tvon forms substantives of state in Indic, just as -thum is used in German. The subjunctive is in part marked by a change of the vowel, as in all languages which follow the old grammar. Agreeing similarly is the formation of the imperfect through change of the vowel in one type of the German verbs. If in another type the imperfect is formed by means of an inserted t, this to be sure is a special characteristic, just as is the b in the Roman imperfect; the principle however is still the same, namely that the secondary determination of the meaning for time and other relationships does not happen through special words or particles added outside the word, but through inner modification of the root.
If, moreover, we add the grammar of the older dialects, of Gothic and Anglo-Saxon for German, of Icelandic for the Scandinavian branch of our language, then we not only find a perfect with an augment, as in Greek and Indic, a dual, more exact gender and relationship markers of the inflections, which now are somewhat worn down and less recognizable; the third persons of the singular and plural of verbs, for example, are complete and in perfect agreement. In a word, in the contemplation of the old monuments of the Germanic language not the slightest doubt can remain that they formerly had a quite similar grammatical structure to that of Greek and Roman.
Even now very many traces of these older forms of language remain in Germanic, in German itself more than in English and the Scandinavian dialects; but if on the whole the principle of the more recent grammar prevails here -- to form conjugation primarily through auxiliary verbs, declension through prepositions -- this should mislead us the less, since also all the Romance languages, which stem from the Latin, have undergone a similar change, as have all the Hindustani dialects, as they are now spoken, which have approximately the same relationship to Sanskrit as the Romance dialects do to Latin. No external cause is necessary either to explain this phenomenon which shows up everywhere the same. The ingenious structure is readily lost through wearing away by common usage, especially in a time of barbarism, either quite gradually, or at times also more suddenly; and the grammar with auxiliaries and prepositions is actually the shortest and most convenient, like an abbreviation for simple, general usage; in fact one could almost establish the general rule that a language is the easier to learn, the more its structure has been simplified and approximated to this abbreviation....
The real essence of this principle of language which prevails in Indic and in all languages derived from it is best made clear through contrast. For not all languages follow this grammar, whose ingenious simplicity we admire in Indic and Greek, and to whose character we tried to call attention in the previous chapter. In many other languages and actually in the most, we find the characteristics and laws of a grammar quite different from that, indeed in complete contrast with it.
Either the secondary markings of meaning are indicated through inner change of the sound of the root, through inflection; or on the other hand always through a separate, added word, which by itself indicates plurality, past, a future obligation or other relationship concepts of manner; and these two very simple cases also designate the two main types of all languages. On closer inspection all other cases are only modifications and secondary types of these two kinds; therefore this contrast includes and completely exhausts the entire sphere of language which is immeasurable and indeterminable with regard to the variety of roots.
A notable example of a language quite without inflection, in which everything that the other languages indicate through inflection is arranged through separate words that have a meaning by themselves, is furnished by Chinese: a language which with its peculiar monosyllabicity, because of this consistency or rather perfect simplicity of structure, is very instructive for the understanding of the entire world of languages.... (to 49-50)
The series of grades of languages, which follow this grammar, is accordingly the following. In Chinese, the particles which designate the secondary marking of meaning are monosyllabic words that exist by themselves and are quite independent of the root. The language of this otherwise refined nation would accordingly stand precisely on the lowest grade; possibly, because its childhood was fixed too early through its extremely ingenious writing system. In Basque and Coptic, as in the American languages, the grammar is formed completely through suffixes and prefixes, which are almost everywhere still easy to distinguish and in part still have meaning by themselves; but the added particles are already beginning to merge and coalesce with the word itself. This is even more the case in Arabic and all related dialects, which to be sure clearly belong to this type in accordance with the greater part of their grammar, while many other things cannot be taken back to it with certainty; here and there we even find an individual agreement with grammar through inflection. Finally, in Celtic some individual traces of grammar through suffixes are found; yet in greater part the newer manner is the prevalent one, of conjugating through auxiliaries and declining through prepositions....
I would really be afraid of tiring and confusing the reader if I reported everything that had been gathered and prepared. Enough if some order has been brought in the whole field and it has been indicated satisfactorily, by what principles a comparative grammar may be drawn up, and a completely historical family-tree -- a true history of the origin of language instead of the former fabricated theories about its origin. What was said here will at least be adequate to demonstrate the importance of the study of Indic, even only from the point of view of the language; in the following book we will contemplate this study in relation to the history of the Oriental spirit.
I conclude with a look back at William Jones, who first brought light into the knowledge of language through the relationship and derivation he demonstrated of Roman, Greek, Germanic and Persian from Indic, and through this into the ancient history of peoples, where previously everything had been dark and confused. When however he wants to extend the relationship to some other cases too, where it is much smaller -- further, to reduce the indeterminably great number of languages to the three main branches of the Indic, the Arabic and the Tatar families -- and finally, after he himself first determined so beautifully the total difference of Arabic and Indic, to derive everything from one common original source simply for the sake of unity; then we have not been able to follow this excellent man in these matters, and in this everyone will unhesitatingly agree who examines the present treatise attentively.