[[Disclaimer: Before I left Berlin for Africa in 1992, Dr. Najock was a colleague of mine at Freie Universität, where we worked together with other colleagues on various projects.]]
Dietmar Najock opens his new book with no fewer than 3 introductions: a Vorwort (p. VII), some Prolegomena (pp. IX and an Einleitung (pp. LII-LIX). The first of this threesome informs the reader about the nature of the book: it is a parergon, composed over several years; the last explains the book’s terminology and methodology. In the middle we find 44 pages of elucidations on the status quaestionis, organised in 10 chapters (titles listed on p. IX) plus an appendix, where 6 classifications of certain ‘Sachgruppen’ (= content-based arrangement by subject word groups) are presented in tabular form.
The most valuable part of these prolegomena is Najock’s survey and discussion of 12 “gebräuchliche Sachgruppensysteme” (systems of subject word groups in use, pp. XII-XIX). Here he offers the reader a masterly presentation of the achievements and the shortcomings of most of the word classifications produced during the last century. His conclusion is that there is no ideal system defying all criticisms (p. XXIII, chapter 5, note 1). Then follows a comparison of the aims as well as the problems of the two leading systems (pp. XIX-XXIII): Franz Dornseiff’s “Der deutsche Wortschatz nach Sachgruppen” (first edition 1934, fifth edition Berlin 1959 and 1970) and Rudolf Hallig/Walther von Wartburg’s “Begriffssysteme für die Lexikographie” (second edition Berlin 1963). In order to avoid creating a new system of his own, Najock decides to follow Dornseiff’s order, since it offers the finer analysis.1 At this point, four groups of problems arise (pp. XXVII-XLIV) and are discussed in detail: human insufficiency; insufficiency of the system; divergencies between language and the system; figurative expressions.
Najock also describes 3 tests (pp. XXI, in which he checked certain features of both systems. Here astonishing inadequacies come to the light: Dornseiff’s system puts ‘summer’ and ‘winter’ into temperature categories (‘warm’ and ‘cold’), but ‘spring’ and ‘fall’ into time categories (‘early’ and ‘late’); clearly this is an uneven classification. In another test 3 doctoral candidates were asked to classify a short Latin text according to the other system. It turned out that only 27% of all cases were analyzed identically by 2 participants and no more than 18% by all three coworkers. The inadequacy of the human factor appears obvious.
The main bulk of the book (p.1-333) contains 7 indices, the last 5 of them indicating frequency (133-299). All of them document each of the 10 Eclogues separately but regularly offer a final summary of all of them. However, this is not yet all: the study is rounded out by 5 chi-square-tests (p.301-308) and 5 binomial tests (309-333), explained in the beginning (p. LV-LIX). What do we learn?
Here the reader must change himself into the user; he will have to study the tables and draw his own conclusions. Najock does not offer results but rather, as his title says, a ‘Schlüssel’, a ‘key’. By using it, the reader may discover thematic gravity centres and prove their existence statistically. While ‘audio’ appears only 6 times in the whole of this corpus, ‘video’ appears 27 times. The user will have to look at the individual verse, at the single word, and take into account its place in the various lists, in order to compare the value of its appearance at a given place. What Najock has given us is indeed a brilliant piece of statistical analysis and a magnificent working tool, and it is to be hoped that many a researcher will use it diligently and frequently.
However, Najock’s penetrating study sometimes appears to me all too sharp. When altus is divided between its meanings as ‘hoch’ and as ‘tief’ (‘high’ and ‘deep’ p. XX) Virgil might have been confused. The poet used only one and the same word, pointing to the vertical dimension; neither he nor his language felt it necessary to differentiate further: why, then, should the modern interpreter do so?
To analyze a language means to follow its own patterns, its proper categories and genuine features. I remember a Chinese student of mine, who asked me about a certain frater in a Ciceronian text. He wanted to know whether this man was the younger or the older brother. When I replied that this differentiation was not indicated in the Latin sentence, he felt unsatisfied because when using his own language he had to choose between one of these two positions. Poor simple Latin speakers, who neglected order and hierarchy, for whom ‘brother’ was simply ‘brother’! This is where the often deplored egestas patrii sermonis comes in, and where the superiority of age-old modern Mandarin makes itself manifest majestically.
I would suggest that each Latin word be treated as one element, however it is used, be it understood metonymice or metaphorice or not. Najock himself is not far from this view when he argues (p. XXVII) that some lemmata, even when used with a specialised meaning, remain very close to the original meaning and might be classified in this and no other way. It appears as a telling detail, that he includes in his list of problems (p. XXIX) such phenomena as “Verwissenschaftlichung und Zergliederung”, i.e., ‘over-scientification and over-structuring’. But to accept or to reject this view and then to act accordingly is the user’s choice: it is she/he alone who will have to decide how far he/she wants to turn this ‘key’.
Notes
1. When I for my part was confronted with this problem in preparing my book on “Our Daily Latin” (“Unser tägliches Latein”, 1992, 6th. edition 2002), I basically adopted Hallig/Wartburg’s system; however, I followed their hint (see Najock p. XVII) that changes might be introduced in order to answer a given need, and added a few new categories. The result seemed satisfactory; it could be easily transferred to the companion book on “Unser tägliches Griechisch” (“Our Daily Greek”, 2nd edition 2002).