Varrone e i Poetica Verba: Studio sul settimo libro del De lingua latina

(I apologize to the BMCR readership for the delay in submitting this review.)

The back blurb of Giorgio Piras’ Varrone e i Poetica Verba announces, “Questo volume cerca di spiegare le motivazioni che hanno spinto Varrone a trattare separamente l’origine delle parole poetiche.” This monograph on ancient rhetorical theory is formidably learned. Piras has riven all the problems he handles to their elements and presents a very thorough study.

The key approach to Varro is stated on page 62, where Piras notes “quanto Varrone ami le schematizzazioni.” The crucial example of this, for the purposes of this study, is at LL VII.5, which reads, “Dicam in hoc libro de verbis quae a poetis sunt posita, primum de locis, dein quae in locis sunt, tertio de temporibus, tum quae cum temporibus sunt coniuncta, sed iis ut quae cum his sint coniuncta, adiungam, et si quid excedit ex hac quadripertitione, tamen in ea ut comprehendam.” This quadripartite structure of locus-corpus-tempus-actio (hereafter l-c-t-a) is a schema; yet says Piras, despite Varro’s love of schematization, book VII of the LL does not conform to the strictest standards of this or any other schematization. Why? The answer to this question is at the heart of Piras’ disagreement with previous scholars of Varro such as Robert Schröter and Hellfried Dahlmann.

Specifically, these previous scholars have tried to read book VII as Varro’s application of the second grade of etymology to words used by Latin poets. This Piras denies. The four grades of etymology are given by Varro at LL V.7-9: “Infimus quo populus etiam venit … Secundus quo grammatica descendit antiqua, quae ostendit quemadmodum quodque poeta finxerit verbum, quod confinxerit, quod declinarit … Tertius gradus, quo philosophia ascendens atque ea quae in consuetudine communi essent aperire coepit … Quartus, ubi est adytum et initia regis: quo si non perveniam ad scientiam, at opinionem aucupabor…” In addition to these two schemata, there is in the LL another, tripartite schema, or more correctly a number of tripartite schemata, since for Varro “triplice è il linguaggio per natura” (127; cf. 47-50). On top of that, there is more than one quadripartite structure to be found (46-47). This situation might present problems also to a student of Varro, and Piras’ argument is in many places very finely drawn. Thus Piras’ frequent references to what he has already said, and what he is going to say, are helpful.

In a brief introduction, Piras presents Varro’s own organization of his book, taken from LL VII.109-110 and V.1. (Piras copiously quotes and summarizes Varro throughout.) Piras here retails arguments from many specialists speculating about the missing pieces of LL. The use of ternary structure “era particolarmente cara a Varrone” (22). Employing this observation, Piras’ gives his chart of the hypothetical whole of the LL (23).

In chapter 1 Piras addresses “La quadripartizione: locus, corpus, tempus, actio.” This structure is found in other works of Varro. In addition to books V and VI of the LL, the incomplete Antiquitates rerum humanarum et antquitatum also had (something very close to) the l-c-t-a structure, according to Augustine, De civ. Dei 6.3 (quoted in full on 28), though Piras must spend some time (26-36) explicating his use of this passage. Piras also demonstrates its use in Varro’s Isagogicus ad Pompeium, Res Rusticae, and Epistula ad Marullum (38-40).

Varro offers a philosophical explanation for the use of this quadripartite arrangement at LL V.11-13, which he assigns to Pythagoras. Piras expands on this (42-45), but also credits Dahlmann’s assertion that what Varro attributes to Pythagoras has in it some Stoic influence (45). The quadripartition of l-c-t-a is found in contexts great and small (46) and in authors other than Varro (50-56), it being “uno schema retorice più generale” (51). What is its source? Aristotle? The Stoics? Varro himself? Piras finds a proximate source in Antiochus of Ascalon who, along with Varro, was something of a philosophical syncretist. “Ebbene Antioco, come è noto, cercò di conciliare le posizioni accademiche con elementi stoici e peripatetici. È quindi possible che egli abbia trasmesso a Varrone uno schema che doveva preesistergli e di cui poi troviamo trace sia in ambito stoico che in ambito peripatetico e retorico” (55). This question of the sources of rhetorical theory exercises Piras throughout.

Piras’ most substantial point, about the relation of book VII to the second of the four grades of etymology, is addressed in chapter two, “Le raggioni del settimo libro.” Piras begins this chapter with a discussion of the other three grades of etymology (57-71), before coming to a long discussion of the second grade (72ff.).

This second grade is concerned with the explanation of the origin of words of poetic origin. Varro’s definition reads in full: “Secundus quo grammatica descendit antiqua, quae ostendit quemadmodum quodque poeta finxerit verbum, quod confinxerit, quod declinarit. Hic Pacui ‘rudentum sibilus,’ hic ‘incurvicervicum pecus,’ hic ‘clamide clupeat bacchium.'”¹ From this passage, Piras discusses the meanings of ‘fingere,’ ‘confingere,’ ‘declinatio,’ and ‘grammatica antiqua’ (74-81). ” Fingere in Varrone serve ad indicare una creazione ex novo” but more narrowly, one based on onomatopoeia, as evidenced by the example of ‘sibilus.’ ‘Confinxerit’ Piras renders as “hanno ‘composto’ nuove parole” (75), again bolstered by Pacuvius’ example. ‘Declinatio’ is “derivazione” (76). Seeking the source of the ‘fingere-confingere-declinatio’ “dottrina delle neoformazioni poetiche” (76), Piras canvasses the Peripatetics, but favors the Stoics, nevertheless conceding that “non sempre è possible dire a proposito dei giudizi varroniani sui suoi predecessori” (79). At LL V.9, Varro mentioned that “non solum ad Aristophanis lucernam, sed etiam ad Cleanthis lucubravi.” Hence, Piras asserts, it’s too fine a question to ask whether by ‘grammatica antiqua’ Varro means Stoic etymology or Alexandrian (80).

But, notes Piras, Varro does not mention this business about neologisms again, which counts against the argument that book VII is an application of this second grade of etymology (81). Rather, reading LL VI.97, VII.5, and VII.110 together, “[E]merge che il contenuto del VII libro è costituito dalle parole poetiche e più precisamente dalle parole poetiche che riguardono la quadripartizione luogo, corpo, tempo, azione” (84). More strongly, it is “evidente non solo che Varrone non ritorna mai sui tre tipi di formazioni linguistiche attribuite ai poeti in 5,7, ma anche che egli non dice mai di aver compreso nel settimo libro i vocaboli ‘creati’ dai poeti, bensì quelli ‘utilizzati’ dai poeti … Infatti troviamo ben poche esempi delle tre categorie del secondo grado” in book VII (87; cf. 111). Other students of Varro have seen the introduction to book V (including V. 7-9) as a program statement, introducing the second triad of books of the LL (book I being an overall introduction), and have tried to fit the contents of book VII into this scheme.

If book VII is not a demonstration of the second grade of etymology, then what is its motivation? The introduction to book VII speaks of the difficulty of explicating poetic language. Words change over time, both in shape and meaning (114). But more than that, poetic language has special qualities which distinguish it from prose and make it harder to investigate. Piras notes among these special qualities the ‘vetustas’ of the words, the ‘auctoritas’ of the poet, and the “nuova ‘consuetudo'” which the ‘auctoritas’ of one writer’s usage can create (116-121). These combine to give poetic language a special liberty. Book VII is about the difficult work of linguistic archaeology (cf. 143) and not about the second grade of etymology.

Chapter 3 is titled “La struttura del settimo libro.” Ternary structure is at the base of LL in many ways (127) as is the quadripartite (128, 133, 178). Refining further a previous statement, Piras notes that the organization of the etymological books of the LL along the lines of l-c-t-a “si basa su categorie che riguardono l’oggetto rapresentato dalle singole parole e non su categorie grammaticali o linguistiche” (128). The structure of book VII, then, owes as much to the words discussed in it as to the l-c-t-a structure. Piras insists that Varro himself had no problem with that: “Non è possible ritenere — come fa Dahlmann — che [VII.5] sia un’esplicita dichiarazione di Varrone della consapevolezza che la quadripartizione stoice è particolarmente violata in questo libro. Forse è vero, come vedremo, che Varrone si allontana di questo schema in maniera particolare in questo libro, ma la possibilita di ‘violare’ questa disposizione è dichiarata dal Reatino nella premessa di tutti e tre i libri della sezione etimologica e in particolare in 5, 13 che fa parte dell’introduzione a tutta la triade” (144; cf. 180).

But, he adds,

“Altra questione è naturalmente quella dell’applicazione practica di questo schema e vedremo come in effetti non è possible riscontrare una aderenza alla disposizione teorica cosi evidente come nei libri precedenti.

“Non vedo comunque motivi per considerare incompatibile teoricamente questo tipo di disposizione con la trattazione delle parole usate dai poeti o che si trovano nei loro scritti.” (144)

In other words, what cannot be proved to work in practice, still should work in theory.

Piras’ burden here is the explication of individual words and sections in book VII (145-178). As we have been warned, poetic words can be difficult to elucidate. One locus “non è chiaro” (157); another “è dubbio” (158); another “è incerto” (159); another has “diversi parti oscuri” (173). In the end, Piras concedes, “Dall’analisi condotta sinora appare chiaro che la struttura del VII libro è più complessa di quella degli altri libri etimologici e presenta molti punti oscuri” (178; cf. 182-183). The incomplete state of the knowledge of Varro’s sources doesn’t help (183-186).

Piras has presented an interpretation of book VII that is less simplistic and more respectful of its author, than his predecessors have done. This approach is surely correct. Varro loves schematization, and Piras has set himself to reveal every possible application of it in book VII. But more than one “problem” in classics could be resolved by unfingering somewhat the grip that generic thinking has on modern scholarship. Could we go further down the path that Piras has pointed? One of his claims is that Varro’s own interests have guided his composition. We might, then, change the emphasis slightly and say that LL VII contains words from poetic sources, which are of interest to Varro, and can be fitted into the l-c-t-a structure, instead of saying that starting from the l-c-t-a structure, the contents of book VII are difficult to explicate. Thus, Varro’s “individual consciousness”² is allowed more space, and the problems of the contents of book VII diminish.

This is a dense, specialist work. I want to emphasize again the immense erudition that Giorgio Piras has brought to bear. The book is filled with subtle readings, some quite speculative (153-154 on ‘vates’), and some more fruitful (161 on the connection of VII 50-57 with Plautus). Overall, Varrone e i Poetica Verba marks an advance in Varronian studies.

Notes

1. Piras addresses the textual problems at 57-59.

2. William Stull, review of Cardauns, Marcus Terentius Varro, BMCR 2002.06.40.