Bryn Mawr Classical Review 2000.11.08
Jürgen Untermann, Wörterbuch des Oskisch-Umbrischen. Heidelberg: Carl Winter, 2000. Pp. 902. ISBN 3-8253-0963-0. DM 168.
Reviewed by George E. Dunkel, Universität Zürich
Word count: 1697 words
The appearance of this work is a cause for rejoicing to anyone who cares about Italic linguistics. Despite its title, it collects the inscriptional remains not just of Oscan and Umbrian, but of the entire "other" (non Latino-Faliscan), or at least of those dialects of which inscriptions survive: Hernican, Marsian, Marrucinian, Paelignian, Paleo-Umbrian, Paleo-Volscian, Presamnitic, South Picene, Volscian and Vestinian.1 The currently fashionable term Sabellian for this other branch is eschewed as pars pro toto (9), though it could not lead to confusion since Sabellian has no inscriptional remains.
By making easily accessible and providing modern, informed,2 unified,3 and responsible commentary on material, some of which has up to now been dispersed through widely scattered publications, this work removes, for a while at least, the study of the Italic dialects from the status of a Geheimwissenschaft in which each researcher must index the entire corpus independently for himself -- a somewhat instructive but very time-wasting duplication of effort. Indeed, the bulk of this volume shows how much work would be needed to do this today, i.e. how impressively the bulk of material (and of scholarship) has grown.
Untermann (henceforth JU) expressly states that he intended to produce a book that not only linguists but also historians and classicists could consult with profit.(9) In this not primarily linguistic forum, a few words should be said about finding items in this book. Due to several explicit editorial decisions, not absolutely all known words of the non-Latino-Faliscan dialects will be found here.
First, only inscriptionally attested material is included; this means that dialectal glosses cited by Latin authors are absent unless they are also found inscribed, so that Oscan meddix (used by Ennius and Livy) is mentioned under meddíss, whereas Sabine fasena (cited by Varro) does not appear anywhere. In fact, dialects such as Sabine which are known only from glosses (Hernican having just recently made the jump to inscriptional attestation) are not represented at all.
Also deliberately excluded are personal, ethnic and local names, on the grounds that these need special treatments of their own.(9) Such a division sounds clean and reasonable, but it will still affect subsequent scholarship by effectively removing possibly relevant material from view. Comparison with the index of E. Vetter, Handbuch der italischen Dialekte I (Heidelberg 1953) will show the tremendously different picture of the material as a whole that this produces, given the high level of onomastic content in these inscriptions. We may be grateful that theonyms at least have not been excluded, presumably because they form less of a system.
Other words for which the reader searches in vain may have been replaced by better readings. Since there is understandably no cross-referencing of outdated readings or ghost-forms, there is no way to find the new interpretation directly, i.e. without already knowing it. Sometimes, one must even know which of the recent interpretations has been accepted: in the case of the flask from Poggio Sommavilla (Vetter Nr. 362), neither Vetter's upothethik:feuos nor Marinetti's pohehike ufs will be found, but rather Rix's poéeí skerfs.4 Even possessing the well-known text-collections of Vetter, Poccetti and Marinetti, after which most inscriptions are cited, will not be enough; only when the consolidated "editio minor", currently being prepared by H. Rix (12; no further information is at the moment available since Carl Winter has no web-site), has appeared will consultation become unproblematic.
A final problem is that of the citation-form. In contrast to glossaries or indices, one expects dictionaries to have a normed system of citation. But here material from various dialects with different alphabets and writing conventions must be presented together; furthermore, the first (or even third) singular present or nominative singular is not always attested in our texts. Verbs are generally cited under the infinitive or third singular present if available in any of the dialects ("to be" is grouped under Oscan ezum, "to bear" under Volscian ferom, "to give" under the Volscian third singular present didet), but otherwise under some form that gives an idea of the present stem ("to go" under the Paelignian second plural eite, "to do" under the Oscan subjunctive fakiiad (despite the availability of the Umbrian infinitive façiu), "to hold" under the Oscan future hafiest). The citation-form of nouns is also liable to be taken from any dialect, not just the one one is looking for -- or even, in the case of (d)iuvepatre, under a form that does not occur at all. For both nouns and verbs, then, one must practically decide for oneself what the stem is and search for that.
The entries include not only the expectable treatments of Belege, Bedeutung, and Etymologie, but also discussions of the Bedeutungsfeld (under anafríss we are referred to diumpaís and aapam, and here further to utur, bio, cisterno, kúru, kaias, and the verb-form kellaked), Wortfamilie, and Wortgeschichte (cf. sub futír). These are a delightful luxury that is easier among Trümmersprachen and show JU's mastery of his material.
The introduction could be short because JU has expressed himself on all the theoretical issues elsewhere, especially in "Zum etymologischen Wörterbuch des Oskisch-Umbrischen," in Das etymologische Wörterbuch, ed. A. Bammesberger, Regensburg 1983, 295-312 and in "Aporien bei oskisch-umbrischen Etymologien," in Oskisch-Umbrisch, Texte und Grammatik, ed. H. Rix, Wiesbaden 1993, 307-325, as well as throughout the "Rest- und Trümmersprachen" discussion (I refer to JU's bibliography in his Festschrift Sprachen und Schriften des antiken Mittelmeerraums, ed. F. Heidermanns, H. Rix and E. Seebold, Innsbruck 1993, 493-512). It is a privilege to be able to compare the practical result in this completed etymologicum with the sample entries of the first and the tortured lucubrations of the latter article. In striking contrast, here JU has gotten himself totally together: choices are made and positions clearly taken; a half-sentence dryly provides the result of such agonies as are depicted there. However, the "freigegeben -- rezeptpflichtig" (over the counter vs. by prescription) classification espoused in 1993 is not applied. Between Proto-Oscan-Umbrian and Indo-European, JU still has problems with the reality of Proto-Italic, which he cannot quite bring himself to believe in as long as he cannot specify its date and location.(10-11)
This work touchingly considers itself (9, 12) to be the third volume of a once-envisioned handbook of the Italic dialects begun by Emil Vetter's text-collection of 1953 (which included, however, Faliscan and dialectal Latin as well); indeed, JU's dissertation of 1954 was the beginning of the present volume. Physically, the difference between the two volumes could not be greater, the first being small and of flimsily bound post-war paper with tiny margins, whereas this large, sturdily sewn volume leaves plenty of space for marginal notes on its acid-free pages -- a healthy typographical reaction to Vetter's crammed pages.
Typography is particularly important in a dictionary and here, the general aesthetic impression is one of software from days gone by. The various sizes of type-face are not proportional; Roman type in reduced size looks squashed out sideways, while Greek is stretched vertically when large. It feels like one is looking at distortion mirrors or wearing the wrong glasses. The bit-map a with macron and acute could not be smoothed (43, 887b).
Also, despite the obvious thought given to formatting for ease of use, the layout is sometimes problematic. Although it is a delight to see the entire Italic material neatly arrayed before one, the presentation of large paradigms (see e.g. ecuc, ezum, izic, or fakiiad) is far from optimal: the glossing comes counter-intuitively only at the end of each group of forms, and the spacing-off of the comments in small type inserts optical divisions at unneeded places. The plethora of abbreviations (subsections, Italian provinces, dialects; why the double hierarchy of abbreviations of Italic dialects ?) slows the reader down considerably without bringing any advantage.
A dictionary offers a refreshingly different perspective on linguistic problems. Released from their sometimes baffling textual surroundings, forms seem once again fresh and challenging. On the other hand, the pan-Sabellian arrays are in fact a different kind of "context," one which can be thought-provoking in a different way. Surfing the index is another good way to learn which long-held comparisons have been invalidated by recent developments in historical Italic linguistics. E.g. Volscian couehriu must be separated from Latin cu:ria < *ko-wir-iya: as deriving not from the doublet *wi:ro- "man" but rather from the verbal root *weg'h- "convey by wheeled vehicle."
Reactions as to the success of the various specific etymological proposals here offered will naturally differ: "[die] Linguisten ... wissen ohnehin alles besser."(JU 1983:295) The particles suffer from the usual disinterest in this morpheme-class: ad is absurdly explained as a "vielleicht alte Sandhi-Variante von *a: und *ati (> *H2eti)."(46) The affixation of an initial laryngeal to each particle (*H2en[e]H2- for *ano ; af-, ap- from "*H2epo"; *H1en for *en [109, 225]) rests on the (unspoken) assumption that particles are structured just like roots, but even if Benveniste's root-hypothesis is correct, which has by no means been proven, this need not at all be the case.5 Bibliography is also an issue here: the discussion of *at (47) ignores Histor. Sprachforschung 101 (1988), 53-78; that of immo (342) ignores Melchert, KZ 98 (1985) 184-205; Lat. enim is taken as an accusative (344), not as the ablative it is (Berthold Delbrück y la sintaxis indoeuropea hoy, hsg. E. Crespo und J.-L. Garcia-Ramon 1997, 63-83). To be fair, though, JU is just reflecting the widespread general attitude toward the fourth type of morpheme.
Despite such minor quibbles (which could easily be multiplied6), we have here an up-to-date life-work by a mature specialist whose care with data and with new theories is unquestionable,7 well aware of the pitfalls of etymology of Trümmersprachen and open about his assumptions -- crucial in a field where etymology is, absurdly enough, often called upon to establish the very meaning of a form (see under Umbrian ekvi!)!
This new standard work is an absolute sine qua non for all future research. In a time when obscurity is sometimes deliberately deployed to impress, Untermann has done something much more impressive: he lays his experience at our disposal and invites us to climb on his shoulders. We owe him thanks for and congratulations on the success of his fifty-year project.
1. The lone Aequian (Ve 226) and Sabine (Ve 227) inscriptions are called Marsian.
2. The bibliography is eclectic; completeness is neither promised nor delivered. Untermann's amici and clientes are openly declared. (8) [ JU mentions the help of co-workers, but the great majority of articles are by himself; cf. however the med- family. There is less of the unevenness that has characterised other collective efforts.
4. For a useful survey of recent new readings of Italic dialect inscriptions see R. Wallace, "Recent research on Sabellian inscriptions," UCLA IE Studies Bulletin 8.1 (1998), 1-9.
5. Oddly, the root *H1ey- "go", which demonstrably did contain a laryngeal, is never so written.
6. A segmentation of the IE optative as *sye:-t, pl. si:-nt (251) disregards the attestation within Latin (sient is older than sint) and the IE allomorphy. Greek πρῶτος does not reflect IE *pro: (582); cf. Doric πρᾶτος. The suffix *-tyo- of Osc. Anaget- implies derivation from a local adverb, suggesting a pre-form *ang'hi-tyo-; and so on.
7. Although note the comparison of Greek λέρας with Latin lapid -- "trotz der Abweichung in Form, Genus und Bedeutung!"(824)