Claudio Vergara presents a theological treatise from Herculaneum, the second in the series La Scuola di Epicuro,[1] from which classicists, scholars of ancient philosophy and papyrologists will surely benefit. Neither the author’s name nor the title of the treatise has come down to us, but there is good reason to believe that Philodemus is its author and that it contains a polemical work against the Stoic doctrine of divine providence.
In addition to PHerc. 1670, Vergara assigns a few more scorze to On Providence (PHerc. 1577/1579, 1636, and 1100) but also excludes a few fragments from PHerc. 1670 (mainly on paleographical grounds). He thus provides the first edition to include all the relevant papyri of On Providence. As a result, the amount of text now available from this treatise has effectively doubled compared with earlier editions. Apart from the attribution of new scorze to the treatise, near-infrared (NIR) imaging and improved methods for reconstructing carbonized papyrus rolls have been exploited by the editor of On Providence leading to many new readings and a better comprehension of its content. NIR images (improperly called multispectral images) are especially useful in the case of PHerc. 1670, which is hard to read with microscope alone.
The first part of the lengthy introduction presents content-based and stylistic arguments which persuasively demonstrate that Philodemus may be the author of the treatise. Following this discussion, Vergara outlines the Stoic view on providence, according to which divinity guides the process of generating and arranging matter. The Epicureans, by contrast, argue against the existence of providence, and in general divine involvement, for good or ill, in the cosmos and human affairs. Vergara then considers the possible Stoic targets of Philodemus’ criticisms, and ends the first part of the introduction with a summary of its contents.
In the second, more technical part of the introduction, Vergara discusses the opening of the papyrus roll, the history of its innermost part, the midollo (PHerc. 1670), as well as each scorza attributed to On Providence, and the disegni of all these papyri. The original roll was around 10.5 metres long, and the text consisted of approximately 3000 stichoi. Of the original 206/207 columns contained in the roll according to the editor, 129/130, that is more than half, are today completely lost. PHerc. 1670 represents a particularly interesting case in the Herculaneum collection, since it is one of the very few papyri which are opisthographs, i.e. contain on their verso text belonging to the same work as that written on the recto.[2] The paleography of the papyrus is also noteworthy: as the author notes, it is unparalleled among the Herculaneum papyri. The main difficulty in the material study of this papyrus roll lies not in its stratigraphy, which proves to be simple enough, but in its reconstruction. The biggest achievement of the editor is the matching of the upper and lower parts of the surviving columns. Although the reconstruction of Herculaneum papyrus rolls can be difficult for non-experts to follow, Vergara explains it in a clear and accessible manner. The outcomes of this reconstruction are nicely visualized in a fold-out page, which accompanies the book. There are, however, a few repetitions in the introduction: for example, the explanation of the numbering of the verso columns with Latin numerals on p. 144 is repeated on p. 151.
After the introduction, Vergara provides tables of concordance. For convenience, it should be noted that cols. 1-162 mainly contain text witnessed by the disegni, along with a few small surviving fragments from the scorze, whereas the text of PHerc. 1670 extends from col. 163 to col. 206/207. The edition concludes with cols. I-IX, preserved on the verso of PHerc. 1670, which are all witnessed only by the disegni and which, as the editor tells us, followed the text written on the recto. The apparatus is rather historical than critical, as it includes conjectures by former editors that are incompatible with the papyrus traces or the width of the gaps.
In what follows, I offer some suggestions on the Greek text, along with a few remarks on methodological issues arising from this edition:
Col. 18: Although the reconstruction of the column seems persuasive, the supplements of ll. 9-11 are highly conjectural and should have been placed in the apparatus.[3]
Col. 35: For the supplement ἀ[λει|τούργ]ητοϲ in ll. 22-23 there is no lemma in the apparatus, so that the reader wonders whose suggestion is.
Col. 37: at l. 3, the first letter of ἀ̣νδ͙{ι}ράϲιν bears an underdot, whereas it should be asterisked (α͙), since the uncertain letter in the disegno is not compatible with an α, contrary to the editor’s claim in the commentary ad loc.
Col. 47: at ll. 28-29 I would suggest εὐϲταθ[(ε)ίαϲ ἀνεπί]|δεκτον. A similar construction occurs in col. 194.25-6.
Col. 62: The letters missing from the beginnings of some of the lines are miscalculated, undermining the reconstruction of the column. For example, at the beginning of line 25, 2 to 2.5 letters are still missing between -[γῶν] and π͙ο͙λὺ.
Col. 78: The phrase κα[ὶ τ]ρε‹ῖ›ϲ … ἀ{ι}πρεπ̣[έϲ]τε[ροι does not make sense and the transcription of the disegno is also mistaken: there we find κα[ ̣₍ ̣₎]ριϲ. The correct reading is probably κἄ[νδ]ρε͙ϲ,[4] which makes good sense with ἀ{ι}πρεπ̣[έϲ]τε[ροι.
Col. 163: at ll. 2-3 I would suggest δίνηϲιν̣ | [τῶν φύϲεω]ν, since ἐ]π̣ὶ̣ (l. 3) is not paleographically possible. Ιn the NIR image, Vergara’s ϲ̣χημάτων cannot be read at all; the inspection of the original does not justify this reading either. The editor should have commented on this issue in the commentary.
Col. 164: at l. 1 the text cannot be verified upon inspection of the original and should not be printed in the text. The commentary on this line stands in contrast to the editor’s decision to print this text. At the beginning of l. 2, π[ρὸ]ϲ should be supplemented. At the end of the line, [καὶ] is highly uncertain and should not be printed in the text.
Col. 174: at ll. 9-10 κ̣ατ’ ἀ̣ρχιτε|[κτονικὴ]ν could offer a more fitting supplement than ἀ̣ρχιτε|[κτονία]ν. The editor could have commented more on ἀ̣ρχιτε|[κτονία]ν, since it is probably attested only once in Philodemus, and all other attestations of the word are very late, with the exception of Galen.
Col. 189: at l. 18 Vergara types ἀρχ̣ή. However, -ρ- is uncertain and should be dotted. (In fact this letter is too slanted to the right to be a ρ.)
Col. 193: It should be noted that the form γεγόνηι̣ in l. 25 is attested only once in Greek literature, and in a poem.
Col. 194: at l. 32 the addition ‹τὸ› of the editor before νοϲεῖ̣[ν at this point of the line, although not impossible, violates Youties’ Law (iuxta lacunam ne mutaveris), and thus should be referred to only in the apparatus. Moreover, the first λ in ἄλ̣|[λωϲ] (ll. 31-2) is uncertain in papyrus, and since the disegno has a ν̣ the editor should have printed α⌈ν̣⌉ at the end of l. 31.
In the same column, the editor does not note that the first letters of ll. 10 and 11 begin earlier than those of the other lines in the column, which cannot be justified by the Law of Maas. Moreover, the letters immediately after δε in l. 10 and τη in l. 11 seem to lie on a sottoposto. Therefore, they cannot be considered in the reconstruction of the text.
Col. 197: at l. 22 the correct reading is not ἀ̣λ̣[λ]ὰ̣, but δ ̣[ ̣ ̣] ̣ (maybe δι̣[ό?). The last letter seems incompatible with α. The following ἔ̣ϲ̣ται̣ (l. 23) is likewise incompatible with the papyrus traces. I read instead the letters ο̣υ̣μ̣-. Consequently, the following supplement μ̣[έ]ν̣το̣[ι should also be expunged from the text.
Col. 200: at l. 25 I propose π[α]ρα⟦ϲ⟧|ϲχών instead of π[α]ρα⟦α⟧|ϲχών based on the inspection of the original papyrus.[5]
Col. 202: at l. 6 the correct reading is ]μερα[ (instead of ]μορα[). I propose ἡ]μέρα.
Col. II: at l. 11 Τ]υδέωϲ, not Τ]ύδεωϲ, should be printed.
Col. VIII: at l. 30 instead of ἤι]δ[ει]ν, which does not fit the space in the disegno ([ει] spat. longius), I propose the aorist εἶ]δ[ε]ν, which perfectly parallels the following ἤκουϲεν (ll. 30-31) and ἔμαθεν̣ (l. 31).
Col. VIII: at l. 4 the letter τ does not match the disegno, and thus should be asterisked. Vergara types ἐπιϲταϲίαν; the disegno has ἐπιϲτ͙αϲία.
At l. 12 a τ is forgotten after ο.
Throughout the Greek text, there are inconsistencies regarding the estimation of gaps and the transcription of the disegni which threaten to undermine the proposed reconstruction and interpretation of the text. For instance, at col. 14.11, the supplement [τ]ὸ[ν λό]γον suggested by Richard Janko does not respect the exact number of missing letters according to the disegno. As Vergara himself notes in the commentary, the interpretation of the period with the supplements provided at l. 11 remains unclear. Under these circumstances, it would have been better to leave [τ]ὸ[ν λό]γον in the apparatus. Furthermore, the participle ὁρί]ζοντεϲ is uncertain: it is possible that ζ should be corrected into τ– something often witnessed in the disegni of this papyrus roll.[6] At col. 193.6-7 the α- of ἀπ[ο]φαϲ|κ̣ομένω̣ν̣ is the result of a correction, but that is not properly indicated either in the text or in the apparatus, which is misleading for the reader. The editor comments on it en passant only in the commentary claiming that before the correction a π had been written; however, I read an υ. At col. 198.4 the letters o and ων in ποι⌈ο⌉υμ[έν]⌈ων⌉ do not correspond to the disegno: ο appears as ϲ in the disegno, while ων is rendered as ϲο. Since the editor has corrected these letters, an asterisk should have been added beneath them to mark this intervention.
Despite these textual and methodological issues, the author has laid a valuable foundation for any further research on On Providence, and this edition will certainly attract the attention of a wide range of scholars in the future. Vergara’s commentary displays eruditio and an excellent command of the secondary literature.
At the end of the book, the reader finds an Index nominum and an Index verborum. Other indices (index of ancient sources and a general index) would have been useful. Finally, the book was well produced, and I did not encounter any typos. I only note one inconsistency regarding Collette 2022 (p. 46 n. 61 and p. 48 n. 66): the bibliography lists this as Collette 2021.
Notes
[1] It follows M. D’Angelo, Filodemo, Opera incerta sugli dèi, Scuola di Epicuro, 20, Napoli 2022.
[2] See K. Fleischer, Philodem, Geschichte der Akademie. Einführung, Ausgabe, Kommentar, Papyri Graecae Herculanenses, 1, Leiden-Boston 2023, p. 81 and n. 1.
[3] Another recent review of this book is G. Roskam, Claudio Vergara (ed.), [Filodemo]: [Sulla provvidenza]. Edizione, traduzione e commento, Méthexis 38.1, 2026, 154-161. Roskam questioned the reconstruction of col. 18, pointing to the uncertainty of several supplements. In particular, he argued that προλέ͙[γειν in l. 5 is doubtful, since the disegno has an ο instead of an ε. However, the interchange between ο and ε is extremely common in the disegni of the Herculaneum papyri. Moreover, although Roskam considers μετ[αβαί]|νειν largely conjectural, one may wonder which alternative infinitive would be more appropriate in this context.
[4] For the frequent occurrence of crasis in On Providence see p. 117.
[5] For the syllabification I propose in the case where a sigma precedes a guttural cf. διδά|ϲκε͙ι͙ν in col. 41.11-12.
[6] Thus, a participle in -τ͙οντεϲ could also be possible.