BMCR 2001.04.06

Sulla preistoria della tradizione poetica italica

, Sulla preistoria della tradizione poetica italica. Firenze: Leo S. Olschki editore, 2000. 179.

1 Responses

Gabriele Costa in his new book Sulla preistoria della tradizione poetica italica (Firenze, 2000) aims to forge a falsifiable scientific hypothesis out of the widely held opinion that some of the early literary/epigraphic remains of the languages of Italy continue the Indo-European poetic tradition (p. 7). One might imagine that an obvious approach towards this perhaps unrealizable goal would be to conduct a careful analysis of the textual evidence itself, and to examine the claims of earlier scholars who have put forth this view for accuracy, coherency and descriptive and explanatory adequacy. In this way we might get a clearer idea of what was wheat and what was chaff and whether the hypothesis could even be maintained. Along the way we might expect new insights to be developed and new analyses to be offered. In fact this is not the path that Costa has chosen. Only one of the three chapters (chapter 2, le vestigia poetiche italiche, pp. 85-116) is actually devoted to a discussion of the evidence and then only in a relatively superficial way. Instead chapter 1 (pp. 7-83) oralità e scrittura nell’italia antica and chapter 3, dalla lingua poetica ai corpora epigrafici (pp. 116-138), contain an incoherent grab bag that seems to me to do little to further the author’s stated goal.

The first chapter contains a brief potted history of the Italic peninsula up to the 5th century BCE and a highly speculative account of the development of the Indo-European poetic tradition from Paleolithic times, which seems to be based on little linguistic data. Costa uncritically adopts a view attributed to Françoise Bader that the Proto-Indo-Europeans were familiar with writing, but that only a few initiates (poets and priests) were privy to this mystery. This remarkable opinion is based chiefly on the fact that the PIE root *peik-/*peig- has in some daughter languages the meaning ‘write’ in addition to the more general meaning ‘paint’ or ‘carve’. But this alone is hardly sufficient evidence to support such a bold hypothesis. The semantic development from ‘carve’ to ‘write’ is so natural that parallel independent specializations cannot be excluded. For example, the Basque nationalist Sabino Arana invented the present-day Basque verb idatsi‘to write’ on the basis of the verb iraatsi ‘to carve’, and many other examples can easily be found.1 An additional argument for this view, also attributed to Bader, is the existence of anagrams and phonic figures in a number of PIE poetic traditions. But these figures, as far as I can tell, need not be giochi alfabetici, as Costa calls them (p. 38), and are perfectly consistent with a purely oral tradition. For example the figure noted by Watkins in Alcman’s (1.36) ἔστι τις θεῶν τίσις perhaps from an earlier * esti kw is t heôn kwitis is perfectly apprehensible by ear.2 To my knowledge a truly alphabetic, visual pun such as an acrostic or the remarkable natu ceu aes of Catullus 60 has yet to be uncovered in any early IE poetic tradition.3

In this volume we are not told much about methodology except that the author follows a “hypothetical-deductive” model (p. 9) combined with liberal use of abduction. Induction, however, seems to be suspect in Costa’s eyes and so it is not surprising that he rarely proceeds in inductive fashion.4 For more methodological discussion Costa refers to his earlier book Le origini della lingua poetica indoeuropea (Firenze, 1998). There we find nothing which I recognize as an explicit discussion of method (e.g. what constitutes legitimate evidence, how alternative hypotheses should be ranked, etc.) but instead a very abstract statement of epistemological principles which seem far removed from the problem of the prehistory of the Italic poetic tradition (for example: la scienza non prova mai nulla; non esiste esperienza oggettiva; la divisione in parti e totalità dell’universo percepito è vantaggiosa e forse necessaria, ma niente determina como ciò debba essere fatto [1998:144] and so on). Methodological explicitness can only be healthy for linguistics, but it seems to me that Costa has pursued his metaworries all the way back to first things and in the process has forgotten to deal with some more immediately relevant problems and questions. One might compare Ranko Matasovic’s recent book A Theory of Textual Reconstruction in Indo-European Linguistics for a much more satisfactory discussion of similar ground.

Let me quote one more paragraph from the first chapter to illustrate Costa’s method of argument (p. 81):

Una bella testimonianza linguistica del passaggio dalla fase vico-paganiga a quella protourbana, è costituita dall’etimologia dell’antichissima divinità romana Vica pota, intesa come il femminile di un * weik(o)poti — ‘signore del * weiko —’ > lat. vicus : “se questa etimologia si rivelasse vera, apparirebbe chiaro il carattere originariamente pre-urbano della divinità”. (A. Carandini, La nascita di Roma, 1997: 208 n. 90).

But an etymology can only be good evidence for something if it is well founded. In fact, the connection of Vica Pota with Skt. vis/pati— ‘leader of the clan’ and Lithuanian vie~spats ‘master’, which originates with C. Hoeing ( AJPh 1903) and which has been revived a number of times, is based upon nothing more solid than Gleichklang. There is not the slightest bit of evidence that Vica Pota had any association with vîcus for the Romans, who indeed clearly connected the name with vinco and potior. And furthermore, there are serious phonological and morphological difficulties involved in deriving Vica Pota from the PIE proto-form *weik(s)potnih2.

In chapter 2 Costa turns to the actual remains of early Italic poetry. He first deals with the by now well-known South Picene texts, whose status as poetry has been clear to the scholarly community for at least a decade. Costa takes the presence of interpuncts between each word in most of the South Picene texts as a sign of their poetic nature and cites Brent Vine’s Studies in Archaic Latin Inscriptions to support this claim (pp. 66, 87). But this seriously misrepresents Vine’s view. For it is not the presence of interpuncts alone but the presence of complex or hierarchical punctuation which may sometimes be used for marking verse divisions.5 Counterexamples of clearly non-poetic texts that mark the division between each word are not hard to find. All of the Tabulae Iguvinae have consistent word-dividing punctuation, even the parts which no one has yet made poetry out of. And further afield the prosaic Old Persian Behistun inscription also uses consistent word dividers.6 Among these he deals with TE. 2, MC. 1, AP. 2 and TE. 5. In his metrical and stylistic analysis he follows closely in the footsteps of the pioneers Watkins and Eichner and does not offer much that is original. There is little in the way of morphological or etymological analysis. For example, in his translation of TE. 2 he translates the very obscure form vepses as sepolto (p. 88) without giving an indication of how he arrived at this translation. Most surprising to me is Costa’s failure to discuss the indigenist vs. Hellenist controversy, for Eichner has expressed the view that the South Picene poems are directly based upon Greek models.7 Watkins, on the other hand, argues for the preservation of an Indo-European tradition in the South Picene texts.8 While I am inclined, both temperamentally and on the basis of the evidence, to agree with Watkins (and Costa) I think that the views of Eichner have been expressed with enough clarity and rigor that they would deserve an explicit refutation in a book such as Costa’s.

On p. 87 fn. 6 Costa endorses the opinion of Marinetti that the Warrior of Capestrano is an instance of an ‘iscrizione parlante’, but this view is based on the analysis of the sequence ma as the accusative of the first person pronoun. Such a form is unparalleled and highly unlikely for an Italic language given the existence of Old Latin mêd, Old Umbrian míom, and Venetic mego. The final chapter seeks to integrate the prehistory of the Italic poetic language with the history of alphabetic writing (p. 116). The main claim presented here is that the Greek alphabet was adopted from Phoenician by poets, since they alone would have had the cognitive tools needed for the initial stages of this process (p. 126). It may or may not be true that poets were instrumental in the adoption and formation of the alphabet, but it is simply not true that they were the only people in eighth century Greece who would have been up to the adoption task intellectually. Whether or not one is convinced by the results of Roger Woodard’s 1997 book Greek Writing From Knossos to Homer, it presents ample evidence that Cypriot scribes used to working with the Cypriot syllabary would have had all the tools necessary to effect the adoption, modification and transfer of the alphabet.

Finally, there are many, many misprints especially in the quotations from English and French scholarship.

Overall, I don’t think this book has appreciably advanced our knowledge or understanding of the prehistory of the Italic poetic tradition. The book may serve a useful purpose if it brings the fascinating remains of early Italic poetry to the attention of scholars who would not normally read a more technical account such as Eichner’s. It may be true, as Costa says in his earlier book, that la scienza non prova mai nulla, but a science (if that is what linguistics is) can at least disprove, and I assume that Costa would agree given his emphasis on the Popperian criterion of falsifiability. Yet time and again Costa has adopted or put forth views that do not stand up to scrutiny.


1. R.L. Trask, Historical Linguistics, 1996: 37.

2. Calvert Watkins, How to Kill a Dragon, 1995: 104.

3. See Peter Bing, The Well-read Muse, 1988 for a discussion of truly visual puns and poetic features.

4. Costa 1998: 138-9 quotes with approval Gregory Bateson’s discussion of induction (here I give the original English): “Many investigators, especially in the behavioral sciences, seem to believe that scientific advance is predominantly inductive and should be inductive… They believe that progress is made by study of the “raw” data, leading to new heuristic concepts. The heuristic concepts are then to be regarded as “working hypotheses” and tested against more data… About fifty years of work in which thousands of clever men have had their share have, in fact, produced a rich crop of several hundred heuristic principles but, alas, scarcely a single principle worthy of place in the list of fundamentals” ( Steps to an Ecology of the Mind, 1972, p. xix). But this statement is hardly a license to ignore induction. Bateson himself says a few pages later: “In contrast I try to teach students…that in scientific research you start from two beginnings, each of which has its own kind of authority: the observations cannot be denied, and the fundamentals must be fitted: You must achieve a sort of pincers maneuver” (pp. xx-xxi).

5. Brent Vine, Studies in Archaic Latin Inscriptions, 1993: 353.

6. Although we should note that no less a scholar than Johannes Friedrich, “Metrische Form der altpersischen Keilschrifttexte,” Orientalistische Literatur-Zeitung, 31, 1928, 238-245, tried to uncover some verse in the Behistun inscription.

7. Cf. Heiner Eichner, “Ein Heldendenkmal der Sabiner mit trochäischem Epigramm eines pikenischen Plautus des fünften Jahrhunderts v. Chr.,” Die Sprache, 34, 1988-90, 98: Die Adaptierung griechischer Vorbilder, die über die Adriaküste nach Italien gelangt sein dürften, an die Erfordernisse italischer Sprachen…sieht man schon im 5. Jh. v. Chr. im vollem Gange

8. Calvert Watkins, How to Kill a Dragon, 1995: 126-134.