chapter  4
Revisions on the basis of new data and new principles
Pages 23

By the early decades of this century Indo-European studies were equipped with large and well-considered presentations on the parent language, the early dialects and the cultural background of their speakers. These were based on theoretical views that had been expressed most forcefully at the beginning of the neogrammarian movement, but, as noted in chapters 1-3, also earlier, at the beginning of the nineteenth century. Indo-Europeanists were concerned with structure, indeed innermost structure. They sought to determine that structure by mastering the data. The data of the early texts were to be treated like information obtained about spoken languages of the present. Respect for data kept them from publishing revisions of hypoth­ eses simply by examining those proposed by their predecessors as bases for intellectual gamesmanship. And when new information was provided, as after the discovery of Tocharian, the Anatolian languages and Mycenaean Greek, the articles and monographs that treated them did not begin with a section on some “theorist” or some specific “theory”; rather, they pre­ sented the data and then proposed their interpretation on the basis of longaccepted procedures and theory. This is not to say that theoretical works were lacking. They were

generally produced by eminent scholars after long concern with many languages. The outstanding Sanskritist, Jakob Wackernagel, produced two volumes on syntax that still ought to be mastered by anyone concerned with this section of grammar (1926-8). The somewhat more daring Havers published an explanatory treatise on syntax (1931). Starting from a differ­ ent tradition, Trubetzkoy laid the groundwork for treatment of phonology (1939). Among those treating methods, Meillet, when invited to give a series of lectures in inaugurating the program of the newly founded Institute for the Comparative Study of Culture in Oslo, presented an excellent exposition of the comparative method in its relationship to typological comparison, illuminating at the same time application of these

procedures to data that takes into consideration the varied social situations in which language is used ([1925] 1967). Meillet’s insistence on dealing with actual speech, like Brugmann’s, and

his pointed statements on the variety of language in accordance with specific social communities, would scarcely have pleased his former teacher, Saussure, nor many linguists today. In one of his crisp sentences he may even exaggerate the differences among varieties within language: “Rien ne differe plus d’un etat de langue qu’un autre etat de langue” (1925: 111). There can be little question that he chose the term langue deliber­ ately a decade after publication of Saussure’s subsequently celebrated series of lectures. The different varieties recognized by Meillet are neither specimens of parole nor of performance, but actual languages maintained by actual language communities. And since the Caucasian languages, rather than his example of varieties within his own French language, were to be the research area of the new institute, Meillet welcomed the pro­ posed undertakings, because they would “provide precise and comprehen­ sive information on states of language, and in this way lead to renovation of theoretical principles [doctrines]’ (1925: 116). We may be grateful that in the country which carried out the most notable political revolution of modern times this term had not been extended to revisions in intellectual approaches. Meillet’s unsurpassed little monograph should be well known by every

linguist, both because of its highly thoughtful conclusions and because of the accuracy of his predictions on the results of detailed concern with specific languages. Some of the most highly regarded linguistic study of recent years was carried out in accordance with Meillet’s insistence that varieties of languages, such as the social registers of English, be studied. Moreover, some of the most fruitful theory of recent times has resulted from examination of the very Caucasian languages to which he referred. The monograph might be characterized as an updated and enlarged version of Brugmann’s manifesto, with more specific attention to actual data than Brugmann allowed himself in his brief statement. Like Brugmann’s state­ ment, it may seem to center on historical study, but the insistence on detailed attention to current French and to investigation of the Caucasian language - among other language - areas indicates that Meillet is dealing with all linguistic study. It was into this highly informed and theory-based linguistic situation that

the new data from Anatolia and Chinese Turkestan were introduced. Linguists have been fortunate that some of their colleagues could read, understand and interpret the data. The early fortunes of the Anatolian materials might virtually be

the subject of a thriller. Their discoverer, Winckler, was reportedly too ill to do much with the texts he excavated at Bogazkoy in 1906; until his death in 1913 his considerate colleagues did not encroach on his prerogative for their publication and interpretation. Thereupon, since the

texts were in cuneiform, Assyriologists undertook to read and interpret them. Interpretation was not hastened by the statement published in 1896 by

the highly regarded classicist, Paul Kretschmer, that no Indo-European language could be expected in the future from Asia Minor. When the young Assyriologist, Friedrich Hrozny, published an article in 1915 pro­ claiming “The solution to the Hittite problem,” followed up by a mono­ graph in 1917 on “The language of the Hittites, its structure and its membership in the Indo-European language family,” the IndoEuropeanists did not believe him. He provided incontrovertible evidence, such as the nominative watar, genitive wetenas “water”; even the presence of a heteroclitic noun comparable to Latin iocur, iocinoris “liver” or cognate with English water, Norwegian vatn did not convince all the sceptics. Further, the social disruptions caused by the First World War were not helpful in bringing about an accurate understanding of the new data. A strange development hampered the potential contributions of one of

the most brilliant linguists of the time, Carl Marstrander. He turned his attention to the language, relying for his data on the brief texts that Hrozny had provided in transcription. Marstrander’s monograph of 1919 included advances beyond Hrozny’s publications, though its adventurous undertak­ ing to interpret the phonology was something of a disaster. After this single publication, Marstrander dropped his concern for Hittite, quite inexplic­ ably. The reason - which we learned from him during personal conver­ sations in 1951 - may provide an insight into at least one apparent irrational development in scholarship. A colleague urged Marstrander to provide some business to an impecunious bookbinder. In impetuous gener­ osity, Marstrander took over all his Hittite materials for binding. That night the bookbinder’s shop burned to the ground. Another early scholar, Forrer, lost credit through imaginative sugges­

tions, such as that the people named Ahhiyava in the texts were Achaeans, a suggestion that after two generations has again found some supporters. Giving up his interest, Forrer turned to Mayan studies in central America, where he was effectively separated from scholarship. It was only when the highly regarded Ferdinand Sommer dealt with the

texts in accordance with the painstaking procedures developed in study of the classical languages that Indo-Europeanists accepted the identification of Hittite as a member of the family. Thereupon capable young scholars, notably Johannes Friedrich and Albrecht Gotze, dedicated themselves to concentration on the language. And a few years later the brilliant young Polish scholar, Jerzy Kurylowicz, published an article identifying some Hittite words that included the sound transliterated by h which he related to Saussure’s coefficients (1927). The implications of his discovery eventu­ ally led to the refashioning of many of the conclusions that had been presented in the standard handbooks, though conservative scholars long

resisted them. Yet after publication of the article, accounting for Hittite and the other Anatolian languages came to be one of the major pursuits of Indo-European study. In the meantime knowledge of the two Tocharian languages had ad­

vanced steadily. The texts were readily understood, being largely trans­ lations of Buddhist works, which were written in the already known Brahmlscript (see 1.6 above). Specialists then needed to edit and interpret them, thereupon to produce grammars and other handbooks. Advances encountered problems that were social and political rather than linguistic; the Tocharian B texts were in Paris, the Tocharian A texts in Berlin; but when eventually published, the material could be used. It was soon clear that the texts were not only recent, from the seventh to the ninth centuries of our era, but also limited in value for comparative Indo-European purposes through heavy modification of the language by languages of agglutinative structure. Although full of intrinsic interest as languages greatly modified in structure and lexicon, Tocharian A and B contributed to Indo-European studies largely through their evidence that disproved the centum : satem classification, besides posing the intriguing problem of how an Indo-European branch came to be located in far-off Turkestan. In somewhat the same way the Mycenaean inscriptions that Michael

Ventris demonstrated to be Greek in 1953 have great intrinsic value for this subbranch, but scarcely provided surprises with reference to Proto-IndoEuropean. One kept hoping for written evidence of laryngeals, not least because of a number of unidentified syllabic signs. Yet the awkward Linear B representation of early Greek, provided in short texts produced chiefly for tax and administrative purposes, furnished only relatively minor contri­ butions, such as evidence for retention of labio-velars as opposed to the p t k of the later language. In short, it was the Anatolian languages that led to revisions of the

proto-language as presented in the standard handbooks. No-one in the academic world can be surprised that these came slowly.