BMCR 2004.12.27

M. Terenti Varronis. Fragmenta omnia quae extant. Pars II: De vita populi Romani libri IV. Bibliotheca Weidmanniana

Marcus Terentius Varro, Marcello Salvadore, M. Terenti Varronis Fragmenta omnia quae extant. Bibliotheca Weidmanniana ; 4. Hildesheim: Olms, 1999-2004. v 1-2.. ISBN 3487108461 €27.80 (pb).

Marcello Salvadore’s [S.] new edition of the De vita populi Romani [DVPR] follows the recent publication of the first volume1 in a continuing project to re-edit the extant fragments of Varro. The four books of the DVPR were last edited in 1939 by Benedetto Riposati in his book-length treatment, and his edition has served as the standard since.2 If this work of Varro’s has been neglected, it is not simply, as S. notes, that the arrival of World War II overshadowed the publication of Riposati’s edition. The DVPR tantalizes with its title echoing Dicaearchus’ βίος Ἑλλάδος and the promise of interesting insights from Varro’s encyclopedic knowledge, but the fragments on the whole disappoint: although there are some fragments of historical and cultural interest, many of them display the etymological tedium which is right at home in Festus, and a substantial part of these are definitions of the dazzling array of jugs available to Romans of the Republic. Nonetheless, the work as a whole is important not only as an example of Varro’s scholarly activity but of Roman cultural history, and any progress in elucidating this intriguing text would be welcome.

S.’s edition consists of a brief introduction, a bibliography and lengthy conspectus editionum, the fragments themselves (each with its own critical apparatus and testimonia), and finally synoptic tables listing concordances with the editions of Riposati, Funaioli, and Kettner. One of the more notable merits of this edition is its cost: it is inexpensively produced with camera-ready copy and a rather cheap paperback binding, but at least libraries (and the occasional individual) need not shell out an exorbitant amount for a slender volume. This rather quick and easy manner of production, however, is mirrored in the contents: the entire volume is vitiated by numerous errors, omissions, and inconsistencies, to such an extent that a reader quickly loses confidence in the text. A compilation of the more serious errors and discussion of individual fragments is appended below.

S. begins the introduction by briefly presenting some hypotheses about why all but a handful of the fragments derive from Nonius but cautiously offers no opinion. After an overview of the scholarly contributions that paved the way for Lindsay’s elucidation of Nonius’ working method, S. devotes most of the introduction to a judicious reconsideration of the Lex Lindsay as a criterion for determining the order of the DVPR fragments. Along the way, he subjects Riposati to criticism for wavering between certainty in Lindsay’s findings and his own “idiosyncratic” ideas about the order of the fragments (a charge highlighted in the bilingual blurb on the back cover).

There are legitimate criticisms to be made against Riposati’s text and arrangement, as there would be of any edition. The vagaries of the “primary” and “secondary” quotations in Nonius, the complexities of the 41 Nonian word-lists theorized by Lindsay, and the usual problems introduced by textual transmission — including the “alphabetical” books of Nonius — all complicate the application of the Lex Lindsay and reduce its usefulness in reconstructing the fragments of the DVPR, especially those outside of Book 1. Any editor thus must arrange the whole of the fragments according to his or her interpretation of them, and S.’s approach is not substantially different from earlier editors: he recognizes the general validity of Lindsay’s law but also wishes to verify whether the Nonian order of the fragments corresponds to an “ordine progressivo reale” based on their “carattere interno.” S., to his credit, places greater emphasis on the shorter sequences to be reconstructed in accordance with the Lex Lindsay and thus restores the order of small groups of fragments separated by Riposati and others. Some of his arguments, such as the use of the attributive “Romanus” as a determinative criterion, are questionable, but his careful analysis of the Lex Lindsay as applied to Varro’s fragments is a valuable contribution to the subject. In addition, his generally more conservative approach to the placement of the fragments, including a category of “fragmenta incertae sedis” not found in other editions, merits consideration.

The introduction has several typographical and formatting errors, but these are minor irritants. In his text of the fragments, however, the errors and omissions become much more problematic. S. has reasonably based both his text and critical apparatus on Lindsay’s edition of Nonius, and while he is to be lauded for printing the entire lemma of Nonius in which Varro is quoted (distinguishing the DVPR quotations by a larger font), he sometimes truncates the lemma immediately following the DVPR fragment, without any indication of this. Unfortunately, he is inconsistent in this practice, and this is just one of many inconsistencies which dog his edition. S. makes a few alterations to the text of the fragments and adds some information in the critical apparatus, but in the majority of cases his text of the lemma as well the critical apparatus reproduce Lindsay’s. Yet both in his reproduction of Lindsay and in his own reporting, he is again inconsistent, even in the typographical conventions used. Clearly an editor needs to make judgments on what to report and how to report it, but it is not clear what, if any, principles guide S.

For example, S. often (but not always) chooses to report conjectures on the entire lemma, even when the bulk of it does not concern the DVPR fragment, but his reporting is extremely selective and haphazard. When a quotation from the Menippean Satires occurs, S. will print Nonius’ text; fair enough, but he will report readings of Cèbe (who is not cited in the conspectus of editions) but not those of Astbury (who is). References to scholars or editions not found in the extensive bibliography are all too prevalent: the relevant information can be found only in Lindsay’s edition. Mechanical reproduction of information found in Lindsay’s edition of Nonius is clearly the source for many of these errors: Mueller’s Nonius pagination, which Lindsay uses in his critical apparatus, often appears, though S. elsewhere uses the pagination of Lindsay’s edition. If such errors were made consistently, the level of frustration would be considerably diminished, but there is no discernible pattern. There are even scribal errors resulting from the transmission of Lindsay’s text: in one instance, S. has committed a classic saut du même au même, with the result that Plautus is made to spout a line from Sallust.

If this careless approach were restricted to portions of the text not from the DVPR, no great harm might come of it, but the text of the DVPR fragments is no better. Most seriously, despite the attention paid to Riposati in the introduction, Riposati’s text is consistently neglected. Riposati made substantial (and often imaginative) additions and alterations to the fragments, occasionally combining separate lemmata, but he was attempting to recreate Varro’s version of Varro, not Nonius’s. There is nothing inherently wrong with S.’s generally more conservative approach in establishing a readable text of the DVPR, but, whatever his judgment of Riposati’s text, it merits reporting, especially as S. more consistently reports readings from Kettner’s earlier edition. The text of the fragments becomes a minefield of uncertainty, and one is left guessing whether the text presented, especially when it deviates from Lindsay’s, reflects S.’s intentions or not.

On at least one occasion, an error seems to have affected the placement of a fragment: S. puts fr. 442 among the “fragmenta libri incerti” and places cruces around the book number “IV” (S. uses IIII elsewhere, but here reproduces the IV that Lindsay adopts in his text). In this lemma, the manuscripts vary between III and IIII, but Nonius quotes the first half of this same fragment in another lemma, where the ascription to Book IV is confronted with no alternative readings. S., however, in quoting this second lemma in his collection of testimonia, erroneously prints “lib. III,” and this apparently leads him to doubt the position of the fragment. Yet there is little question that this fragment belongs to Book 4, as in the editions of Kettner and Riposati.

The testimonia following each fragment also present a confusing variety of material. One would expect the passages adduced to reflect or clarify the fragment at hand. However, S. often offers passages which seem to have little to do with Varro’s words or alleged meaning. Sometimes, he quotes passages that give further examples of the Nonius lemma rather than the DVPR fragment. For example, fr. 418 is quoted by Nonius because of the plural use of “pax”: S. quotes several passages, from Lucretius to Augustine, whose only connection to Varro’s fragment or one another, is the occurrence of “paces,” and it is unclear what the testimonia are meant to illustrate aside from Nonius’s point that the form is relatively rare. It is better that S. err on the side of completeness in his testimonia, but often truly apposite testimonia are drowned in a sea of needless repetition. In one case, he quotes six medieval glossaries which all contain some minor variation of “cernuus in caput ruens (in the DVPR fragment, Varro uses the verb “cernuebant”). Much as with his handling of the text of the fragments, it is not clear what has guided S.’s selection, and one is never sure what kind of testimonia to expect.

While the introduction and the arrangement of the fragments deserves serious consideration, the text of the fragments does not. The carelessness and sloppiness not only erode confidence in the text of this volume but also do not augur well for future volumes. It is not a pleasant task to write such a negative review; S.’s task is a complicated and difficult one, but at the very least this edition could have used a proofreader. This book was published by a respected press, and it is puzzling how the editors let so many serious mistakes find their way into print. Lastly, S. himself is particularly ungracious to previous scholars: aside from his treatment of Riposati, S. chides Mommsen for a conjecture by citing it as “inepte” and soon thereafter we find “stulte Manzo.” Mommsen did little, if anything, that could be described as silly (his conjecture here is anything but), and though I am not familiar with the judgment of Manzo to which S. refers, the comment reeks of personal animosity rather than scholarly disagreement. Such words condemn rather S.’s own judgment, in evidence throughout this disappointing edition, which is to be used, if at all, with extreme caution.

Comments on Individual Fragments

fr. 284: S. in the app. crit. says “post fictum lac. signavi” but does not indicate any in the text of the fragment; whether this is from prudence or error, I cannot tell. His proposed supplement makes little sense in light of the preceding “sed.” S. also describes Mommsen’s conjecture “Panda” as “inepte,” but when he discusses the fragment in the introduction (p16), he seems to assume that Panda is the goddess meant by the transmitted reading “hanc deam.” The etymology given in the fragment for “hanc deam” is “a pane dando”; Panda is a rarely attested goddess, but though S. may have legitimate grounds for rejecting Mommsen’s conjecture, it seems incomprehensible that it could be characterized as foolish. S. also misprints Riposati’s supplement by reversing the order of the last words.

fr. 291: The important variant “ut olim reges” for “ut non reges”, found in some mss. and in Brunetti, is not reported.

fr. 293: S. has omitted two lines of Lindsay’s critical apparatus which contain much important information on the fragment. Most importantly, the reading “utique omnia regis” printed in S.’s text is Lindsay’s emendation (mss. read “in quae o. r.” or “in qua o. r.”). S. reports Mueller’s conjecture “ regis”, but omits Mueller’s addition of “namque”; likewise, Scaliger’s reading is reported as “regi[i]s,” but in full he proposed “quia omnia regiis.”

fr. 298: S. reports that Quicherat corrected the mss.’ “incerni” to “incernere” (as in Lindsay), but does not mention that Riposati retains “incernere.” Riposati’s change of “solent” to “solere,” following Funaioli, is not reported.

fr. 300: The variant “quot” for “quod” is not reported.

fr. 301: Riposati thinks this fragment a Nonian paraphrase, comparing Servius on Aen. 8.479 and Servius Dan. at Aen. 10.164, which S. should have quoted in the testimonia.

fr. 306: Read “plumea” for “plumes.”

fr. 313: S. reports Kettner’s “novum doliis promptum,” but omits Riposati’s reading “novum doliis prolatum.”

fr. 315: Wessner’s lengthy supplement, accepted by Riposati, is not reported.

fr. 316: S.’s text places “tu autem murmurina” in cruces, but within the cruces places brackets so: mur[mu]rina. Although S. hesitantly suggests a lacuna before “tu,” he does not report Riposati’s “tam vinum murrinam” (which Riposati attaches to the end of the previous fragment), nor his addition to the end of the fragment, [eoque aquam addidissent].

fr. 318: Mueller’s suspected lacuna after “aduri” is not reported.

fr. 321: “quod est poculi genus” was deleted in the editions of Popma, Kettner, and Riposati (the latter not reported); S. retains it, and though the phrase may very well belong to Nonius, it seems highly improbable that the phrase belongs to Varro, especially following the words “pocula quae vocant capulas ac capides.”

fr. 322: Read “oris longi” for “oris longis.” The app. crit. reads “mori Aa,” but should read “mori LBaCa: in mori Aa” (apparently a scribal error).

fr. 323: Read “guti, sextarii” for “guti sextarii” (in a long list of wine jars separated by commas).

fr. 324: In this fragment, from Scholia Veronensia in ecl. 7.33, S. prints “tria enim haec similia [sunt]” (his deletion). However, “enim” appears not in this scholium, but in Servius Dan., which reads only “tria enim.” Thilo supplemented the latter with on the basis of the Veronese scholium, and S. has apparently conflated the two.

fr. 326: S., against other editors, attributes the final rhetorical question to Nonius, though the use of the bare “antiquissimos” without “Romanos” violates a (questionable) principle he sets forth in the introduction. More importantly, it seems improbable that Nonius in composing his dictionary would rhetorically ask “ita apud antiquissimos manale sacrum vocari quis non noverit?”

fr. 330: S.’s eye has gone astray in reading Lindsay’s critical apparatus: he omits five lines and confuses the rest. He reports a correction “castulam” for a Plautus quotation, but it belongs to the Varro quotation. Also, for “castulam” in line 5 read “castula.”

fr. 333: S. omits entirely the variants for the DVPR quotation, notably the opening preposition “ex,” which S. prints but is Iunius’ emendation (“et” is the transmitted reading).

fr. 335: “saviunt” is a correction by Mercerus.

fr. 341-377: These fragments, all from Festus, constitute an appendix to Book 1, entitled “Fragmenta vel falsa vel ad hunc librum temere adscripta,” but are not to be found in the any previous editions. The guilty party here is named as Willers, whose name is mentioned nowhere else. Presumably, the reference is to Heinrich Willers’ De Verrio Flacco glossarum interprete disputatio critica (1898).

fr. 379: The words “Varro de vita” have been transposed with “sibi in collum inponit” in the line below. S. prints “lib. I”, but, as this is the second fragment of Book II, he apparently accepts the “II” found in Popma. The readings reported in the critical apparatus are erroneous in the extreme.

fr. 380: S. omits three lines from the middle of the lemma, containing a second quotation from Accius.

fr. 387: There is no critical apparatus, and thus Mueller’s “eidem erat” for “id erat” and Lindsay’s conjecture “derat” are not reported.

fr. 390: The critical apparatus is again missing. Read “obstetrice” for “ostetrice.”

fr. 394: Riposati’s addition is not reported.

fr. 395-396: La Penna believed these fragments to refer to the destruction of Carthage; S. rejects this view on the grounds that, although Nonius’ cites both from Book II, Varro does not discuss the Punic Wars until Book III. He therefore suggests the fragment may belong to Rome’s subjection of her neighbors following the regal period. He does not mention that Riposati, without referring them to a specific era, places them immediately after fragments that he believes (somewhat imaginatively) refer to victory over the Samnites and Pyrrhus.

fr. 400: S. prints “lingula” without indicating that this is Quicherat’s correction for “lingua.”

fr. 407: Read “adhibetur” for “-entur.”

fr. 408: S. has committed a particularly egregious scribal error, jumping from the occurrence of “senati” in a quotation from Plautus quotation to one from Sallust, with the result that Plautus is made to speak rather prosaically of Q. Marcius Rex Faesulas and Q. Metellus Creticus (read “Rex” for “rex”). “Senatuis,” the final word of the Varro quotation (following a lacuna proposed by Mueller) is not found in the manuscripts of Nonius and was not printed by Lindsay. It was an addition by Kettner, but this is not reported by S.

fr. 411: Read “terram” for “terra.”

fr. 423: S. fails to report that Mueller and Riposati attribute the remainder of the lemma after “quinque” to Nonius.

fr. 426: Read “duorum” for “durum.”

fr. 427: Read “corr. Lipsius” for “corr. Lachmann.” Mueller’s lacuna after “Magnum” is not reported.

fr. 432: S. reports that this fragment is lacking in the Popma, Durdrecht, Bipontina, and Brunetti editions, but this is not entirely accurate. (These latter three editions are cited throughout the work by abbreviations which are explained not in this volume, but in S.’s first volume.) A much longer quotation that ends with the same brief sentence Nonius quotes is found in Pliny’s Natural History, so all these editions, as well as Kettner and Riposati, sensibly use Pliny as the source for the fragment. S. for some reason relegates this longer quotation to his testimonia.

fr. 433: Read “Verrem” for “verrem.” In the testimonia, S. quotes several passages where “carbasina” is used as a synonym for “vela” (in sense of curtain), but ignores the possibility Varro is here talking of clothing, as the two Vergil passages Nonius quotes attest. This possibility is strengthened by “ut pellibus,” which seems unlikely to refer to curtains.

fr. 434: Read “miliis” for “militibus” and “adipiscantur” for “-atur.”

fr. 435: Among the numerous conjectures, S. does not include Riposati’s “aut dominatus quos appeterent.”


1. Salvadore, Marcello (ed.) M. Terenti Varronis. Fragmenta omnia quae extant. Pars I: Supplementum. Bibliotheca Weidmanniana. Hildesheim: Olms-Weidmann, 2004.

2. Riposati, Benedetto. M. Terenti Varronis De vita populi Romani. Fonti-Esegesi-Edizione critica dei frammenti. Milan: Società Editrice Vita e Pensiero, 1939.