Nineteenth-Century Romantic Linguistics: The Tree and the Family
Nineteenth-century historical linguistics establishedâand was obsessed byâthe idea of language âfamilies.â Unlike the eighteenth-century Enlightenment concern with spatial arrangement and classification, nineteenth-century intellectuals were concerned with time and development. As positivist progressives living in the age of capitalist expansion across the world, they believed in increase and ramification in all things, and, as romantics and geographical determinists, they liked to see simple roots nourished by native soils. Thus, the image of a tree growing and spreading through the ages became dominant in the development of species, or âraces,â and languages. Above all, they insisted that good languages were organic, growing from the inside, not inorganic, imposed from the outside.1
The standard linguistic terms âloanâ and âborrowingâ themselves indicate something interesting and important about the early nineteenth-century romantic scholars who worked out the Indo-European language family. To them, such terms suggested impermanence and distasteful âtrade.â âFamilyâ and âgenetic relationship,â on the other hand, suggested permanence and propriety. Similarly, the model of a ramifying tree was not only aesthetically attractive and satisfying but it was also able to explain many relationships between and among languages.
This idea leads to the general principle upon which my whole project is organized: there are no simple origins. Thus, imaging historical linguistic or biological development by using the model of a tree, with a single stem from which grow branches and ever smaller branches and twigs, is seldom useful. Only if one takes the multiple roots into account can the tree model sometimes be useful. The past, I believe, is better envisioned as a river in which currents come together to form a unity, then diverge and combine with others to form new unities and so on. The uncertainty of this image should not lead to despair or paralysis. The fact that the chase is endless adds to, rather than detracts from, its fascination.
Early perceptions of relations between languages
Language families were envisioned long before the nineteenth century. Probably even under the Assyrian and Babylonian empires Jews were aware of the obvious relationship between Hebrew and the official language Aramaic. Jews living within Islam added Arabic to the cluster. Judah Halevi (1075â1141 CE) the Andalusian poet who wrote in Arabic and Hebrew, was quite explicit on this relationship. He maintained, in the orthodox way, that Hebrew was the language of God and should, therefore, be used in prayers. He went on to the unorthodox notion that Abraham had spoken Aramaic in everyday life and had taught it to his son Ishmael who had then developed Arabic.2 There was a difficulty here in that Hebrew was seen as the paternal language, whereas Arabic is, in fact, much more archaic. Only after the weakening of Judeo-Christian influence on language studies at the end of the nineteenth century was Arabic seen as closer to the original Proto-Semitic than Hebrew.
In the sixteenth and seventeenth centuries, when Portugal became involved in Ethiopia and the Ethiopian church contacted Rome, the Ethiopian classical liturgical language of GeÊŒez was added to the cluster of Hebrew, Aramaic and Arabic.3 The outer bounds of the family began to be defined by the AbbĂ© Jean Jacques BarthĂ©lemy in the 1760s, when he argued plausibly that, although there were similarities between Coptic and Semitic languages, the Egyptian language did not belong to what he called the âPhoenicianâ language family.4 In 1781 the Göttingen scholar A. L. Schlötzer gave the scheme academic ratification as the âSemiticââfrom Noahâs son Shemâlanguage family.5
Earlier in the eighteenth century, a number of other language âfamiliesâ had been recognized. Malay, Malagasy and Polynesian were seen as related in the family now known as Austronesian. The Uralic family, including Finnish and Hungarian, was also established.6
Most historical linguists consider these discoveries unimportant. For them the crucial step was what is seen as the heroic foundation of Indo-European studies: William Jonesâs Third Anniversary Discourse to the Royal Asiatic Society at Calcutta in 1786. Jones set out what he saw as the excellence of Sanskrit over its âsister languagesâ Latin and Greek. He saw all these languages as coming from a common source. In addition, Persian and, with some alterations, both âGothickâ and âCeltickâ also derived from this source.7 Like Schlötzerâs establishment of Semitic, Jonesâs scheme, too, had antecedents. Medieval and Renaissance scholars had long recognized that in some languages God was called variants of Deus, in others of Gott and still others of Bog, thus defining the Romance, Germanic and Slavic families. By the eighteenth century these three families had been quite thoroughly worked out.
As early as the sixteenth century, European priests and other travelers had noticed similarities between Sanskrit and European languages.8 Where Jones went beyond his predecessors was in focusing on similarities of morphologyâthe conjugational systems of verbs and the declensions of nouns and especially common irregularitiesârather than on mere resemblances between words.
This emphasis has worked well for closely related languages. Furthermore, to study morphology one needs a thorough training, which benefits the guild of professional linguists by keeping out amateurs. This is an important factor behind the preference of historical linguists for morphology over lexicon.
The insistence on the relative unimportance of lexical parallels has hampered attempts to connect more distantly related languages. In fact, common words often outlive morphological parallels. For instance, we know from linguistic history that Russian and English are both members of the Indo-European family. Today, however, the two languages show no morphological parallels. On the other hand, a number of basic words in Russian and English, those for âmother,â âbrother,â âson,â âmilkâ etc. are clearly cognate. Thus, lexical comparison is still an essential tool for relating languages, despite potential confusion caused by chance and the possibility of loaning.
The legend of âscientificâ linguistics
By the last quarter of the nineteenth century, linguistics had become a well-established academic discipline. As such it required a genealogy. Therefore, German and Scandinavian practitioners established a standard historiography, or hagiography, of the development of âscientificâ historical linguistics. According to their scheme, the discipline passed through four stages or generations. The precursers were Sir William Jones and Friedrich Schlegel; the founders were Franz Bopp, Jacob Grimm and Wilhelm von Humboldt. The consolidators included Georg Curtius, August Schleicher and August Fick and the final developers were the Junggrammatiker or Neo-Grammarians, August Leskien, Karl Brugmann, Hermann Osthoff and Berthold Delbruck.9 The list reveals a number of interesting and significant features. First, it consists almost entirely of German men. Apart from Jones, only two other non-Germans are referred to as having played roles that do not fit well in the sequence: the Dane Rasmus Rask and the Italian Graziado Isaia Ascoli. The last is the only scholar whose native language was not Teutonic.
In the late twentieth century, the historian of linguistics Hans Aarsleff set up a broader view of the disciplineâs origins. Aarsleff argues that the central figure Wilhelm von Humboldt not only spent his formative years in Paris but also drew heavily on the linguistic ideas of French figures of the Enlightenment, especially Ettienne de Condillac, Denis Diderot and Joseph-Marie DegĂ©rando.10 The failure to note this influence reveals a desire on the part of the nineteenth-century historians to portray âscientific linguisticsâ as an essentially Germanic discipline. It also diminishes the role of gentlemen scholars in its development to the benefit of professional academics.
The historical linguist and historian of linguistics Anna Morpurgo-Davies has described Humboldt as âembarrassingâ in two major respects: First, his intermediate relationship between the Enlightenment and the romantic positivists of the nineteenth century and, second, the ambiguity of his position between amateur and professional.11 The second embarrassment is true of the founders of all academic disciplines. The first is the more interesting. In Volume I of this series, I sketched out the central role of historical linguistics in the formation of the modern university in Germany and elsewhere in Europe and North America.12 I focused on Humboldt, whom I portrayed as the founder both of the Prussianâlater Germanâacademic system and of modern romantic-positivist linguistics. This view must now be modified in view of Aarsleffâs work. Humboldt clearly remained part of the Enlightenment in his broad interest in all languages and in his concern with both their diachronic-historical and their synchronic structural aspects. In these respects he was very different from his successors or one might almost say, in some cases, his products. The latter were exclusively concerned with historical linguistics and with the Indo-European language family and very largely with Germanic or classical languages.
Humboldt was, however, a romantic in his conviction that inflected Indo-European languages were ineffably superior to all others. For him, Sanskrit was the perfect language and Greek the most harmonious.13 Sanskrit was knocked from its pedestal as the original mother tongue by the professionals in the second half of the nineteenth century, but throughout the twentieth century Greek maintained its unsullied reputation as the most harmonious language in Indo-European and, hence, world languages. The only challenger was Latin, which had preserved more of the nominal cases of PIE than Greek. Nevertheless, despite the German and British identification with Rome, consolidated after 1870 when both claimed to be empires of the Roman type, academics tended to prefer Greek.
Jonesâs immediate successors were less cautious than he in their family scheme. Where Jones saw Sanskrit, Greek and Latin as âsisterâ languages descended from a lost parent, German scholars of the mid-nineteenth century tended to see Sanskrit itself as the ancestral language. Only in the 1860s did scholars begin to see that, although the ancient Indian language was archaic in many respects, in others it had made more innovations. Thus, a trend emerged to revert to Jonesâs position and give equal or even superior status to Greek and Latin. It should be noted at this point that German historical linguistics did not fit the general progressivism of the nineteenth century and that for the linguists preservation of original features was considered the mark of a superior language. This value influenced German scholars to move away from the name IndoeuropĂ€isch âIndo-European,â proposed by Bopp to Indogermanisch.14 The suggestion was that the Indo-Aryans and the Germans had been the last to leave their supposed Central Asian homeland and had, therefore, preserved the purest form of PIE.
The Neo-Grammarians
The scholars who made the decisive shift away from Sanskrit as the mother language were the Neo-Grammarians. At one level the move was forced by the recognition that the vowel system of the European branches was objectively more archaic than that of Sanskrit. On another level this shift was an assumption of European superiority in the decades of imperial or more specifically of GermanicâGerman and Englishâtriumph over the globe and the final jettisoning of all Oriental influences upon Europe.15
The Neo-Grammarians were described as the final developers of nineteenth-century historical linguistics. Mostly based in Leipzig, they flourished between 1870 and 1900. The epithets ânewâ or âyoungâ attached to branches of academic disciplines usually indicate continuity rather than the break they wish to indicate. Such continuity was certainly the case with the Neo-Grammarians, who basically only ratified or fixed previous trends. In the previous decades historical linguists had begun to question the maternal role of Sanskrit. Although the Neo-Grammarians claimed to have broken away from their predecessorsâ âorganicismâ (the belief that language was an organism with a life of its own independent of th...