This is the first comprehensive and commented critical edition of the extant testimonia and fragments of Sappho produced so far; it follows a previous, noteworthy collection edited in 2017 (Saffo. Poesie, frammenti e testimonianze, in joint authorship with F. Cinti), which made Sappho's oeuvre accessible at a more affordable price. The new volume provides an introductory section covering several key topics, such as Sappho's life, transmission, ancient editions, Nachleben, language and style. Then there are the Greek text and an Italian translation, followed by the commentary, a thorough bibliography, a conspectus and indexes. The major merits of this impressive achievement are immediately obvious: (i) N. includes not only some astonishing, albeit controversial, recent papyrus finds such as the ‘Newest Sappho’ (with due attention to its fraudulent provenance: cf. B. Nongbri, https://bmcr.brynmawr.edu/2022/2022.05.2), but also several small fragments omitted by E. Lobel / D. Page and E.-M. Voigt; (ii) he offers an updated and sound Italian commentary by reviewing a significant amount of bibliographical items spanning many centuries of scholarly works; (iii) he also provides the fullest collection of Byzantine passages (F 286A–M) assembled so far, which document Sappho's reception after late antiquity. The apparatus is tripartite: sections on metres and sources (including useful indications on the editorial history of each item) are followed by the apparatus criticus (equipped with plenty of details on papyri readings, modern conjectures etc.).

That there will be some points of disagreement in such a massive book is not surprising: (a) N. preserves Voigt's numbering with the addition of alphabetic letters for new items; but this choice is a little awkward, especially since the latest evidence provided by the ‘Newest Sappho’ allows us to reconstruct the poems’ order in Book 1 quite differently from previous editors (we can now follow a ten-poem-long sequence as arranged in the Alexandrian edition). Likewise, the choice to gather sources on the ancient edition (organised into nine books?) in a dedicated section within the testimonia (F 226–35) at the end does not seem appealing. Unlike Lobel–Page, we have no obvious indication of the book division throughout the entire sequence of fragments (just as in Voigt's edition). It would have been more appropriate to have the sources mentioning Sappho's specific books printed as subheadings before the fragments themselves: if the main goal of an editor of Greek lyric poems is to comply with the Alexandrian editions, a clear-cut arrangement of the fragments into books, if known, would have been rewarding. (b) In some cases N. limits himself to setting out the findings of previous or more recent scholarship: for instance, if M. de Kreij (CPh 117 [2022]) is right in assuming that the asteriskos in fr. 44A (assuming that the ascription to Sappho is tenable) implies that the metrical sequence is uneven, the item should be removed from Book 2 and printed among the frr. incertae sedis. N. seems to adhere to Lobel–Page's view, who envisaged traces of marginal notes, but a reappraisal of a digital image of P.Fouad 239 seems to confirm de Kreij's conclusion; this crucial point should have been discussed in more detail. (c) It would have been wise to provide a concise overview of the papyrological transmission in the introduction (at p. 72 n. 400 readers are referred to the Cambridge Companion to Sappho [2021]) as well as of the ancient grammatical sources: in particular, among the scanty remains of Aeolic dialect scattered in the Byzantine treatises on the dialects (some of them attributed to John Grammaticus, maybe John Philoponos), we find words such as cτρότοc (cf. F 16.1) and θροcέωc (Περὶ Αἰολίδοϲ III §.1, p. 215 Hoffmann) that might directly derive from Sappho or Alcaeus. (d) The choice to print the fragments (cum verbis, sine verbis and the glossemata) separately from the testimonia (avoided by Voigt) implies that some items are printed twice (both equipped with the same extensive apparatus): cf. for example F 168E(1–2) = 214B, fr. 2.

Quibbles and slips are understandable in such a monumental volume (N. has since offered several corrections and addenda in Eikasmos 33 [2022], 1–15). Some passages deserve comment: introduction, p. 11: Aristotle's school (his pupil Callias of Mytilene worked on Sappho and Alcaeus) might also have influenced the textual arrangement with respect to some graphic peculiarities (cf. e.g. the replacement of intervocalic ζ with ϲδ with the discussion in Aristot. Metaph. 1.9.35, which is likely to imply direct knowledge of current Lesbian pronunciation): inconsistencies in pre-Alexandrian papyri such as P.Köln inv. 21351+21376 can instead reflect a different textual strand (influenced by Athenian reperformances?). p. 76: a lesser-known and bizarre clue to Sappho's relevance for Humanists is her presence in the ideal library of Cardinal Bessarion's long-time secretary Gaspare Zacchi (item nr. 105 Sappho lirica in the book inventory transcribed by F. Di Benedetto, Miscellanea di Studi in memoria di A. Saitta Revignas [1978], p. 187). F 3: the textual relationship between P.Berol. 5006 and P.Oxy. 424 should have been noted by means of half-squared brackets, as elsewhere (see F 5). F 6(b): l. 9 should be printed and aligned rightwards. F 29(20): l. 6 δαι ζαφ[: a form of ζᾰφεγγής or ζᾰφελής (cf. Hesych. ζ 73-4 L.-C.) next to ζαφ[οίτ- (suggested by B. Snell in comparison with F 96.15)? F 44 (comm. p. 639): ποτήριον is not necessarily Aeolic, if it occurs also in CEG 454 (‘Nestor's cup’: last quarter of the eighth century bce); p. 641: the form Περάμοιο could be better explained as a compromise between the expected Aeolic Περράμω (cf. EM 665, 40 G. [from Herodianus’ On modifications of words] Πέρραμοc⋅ Πρίαμοc Αἰολικῶc) and the metrically suitable Homeric Πριάμοιο. F 46: according to the new edition of Herodian's On Anomalous Words by A. Papazeti (2008) the fifteenth-century MS Hauniensis 1965 has επιμάλθακαν. F 61–53.71.87: N. rightly advises caution in accepting Puglia–Ferrari's reconstruction of several scraps from P.Oxy. 1787+2166(d)+P.Halle 3, given that the physical joints of some scraps need to be checked against the original and that the entire issue concerning the textual arrangement of these pieces and their relationship to Sappho's fourth book is still in need of reappraisal. F 103: besides [Ammonius] also Herenn. Phil. De div. verb. 33 Palmieri should be quoted. F 103C(4): l. 1 should be aligned with the beginning of l. 2. F 148: sources (IV) and (V) ought to be deleted: [Plut.] Nob. 5 and its Latin ‘translation’ are both sixteenth-century forgeries; the reference to Sappho is likely drawn from Calliergi's edition of Pindar scholia (1515), as suggested by S. Boscherini (Tradizione classica e letteratura umanistica. Per Alessandro Perosa [1985], p. 657; see now also D. Blanck, in: J. Martínez [ed.], Fakes and Forgers of Classical Literature [2011], pp. 33–59). F 161: there is blank space in P.Bouriant 8, col. vii before καί, thus probably F. Meister, ZPE 214 (2020), is right in assuming that the quotation starts there. F 168B: on the text see D. Clay, TAPA 101 (1970), 119–29 and P. Reiner / D. Kovacs, Mnemosyne 46 (1993), 145–59. F 186: John Philoponos’ treatise should be quoted according to G. Xenis's recent edition (2015). F ˚223C: P.Oxy. 2310, col. i, 8–13 quotes Sappho twice (the reading is possible in both places). Testimonia: in P.Vindob. 39966v, col. i, 7 (a book-list dated to around the middle of the first century ce) Σα]π̣φ̣οῦς has been tentatively suggested by E. Puglia, ZPE 123 (1998), 81, and can be added to this section. F ˚287: the text is usually quoted according to the MS Leiden, Voss. gr. 20 (siglum T) collated by R. Reitzenstein, but another witness of the same Byzantine treatise (likely derived from Herodianus) is the hitherto unexplored Vindob. phil. gr. 254 (end of thirteenth century, contemporary with T), in which the fragment appears without any relevant variant. F ˚297: I wonder whether both scraps contain Aeolic poetry: (i) the hand of (b) seems to be different from (a): cf. my and epsilon (correctly noted in the ed. pr.); (ii) at (3).2 ]το̣ νόημα looks suspicious, since in Aeolic we would expect νόημμα (a few inconsistencies in the paradosis are, however, attested: see F 60.3–4 νόημμα and κάλημι).

N. is to be congratulated for this brilliant and well-produced work of rigorous scholarship, which will undoubtedly serve as a fundamental resource for anyone interested in the fascinating remnants of Sappho.