Nec, siquid olim lusit Anacreon, | delevit aetas (Hor. Carm. 4.9.9–10). In the battle against the ravages of time, B.'s Anacreontic commentary represents a worthy weapon. It is apparent from the preface that the focus of the book is on the commentary, which indeed occupies pp. 255–861 of the work. This is further demonstrated by the fact that B. (a) mostly – although not slavishly – keeps to Page's selections and choices for the melic and iambic fragments (the equivalence of the latter with West, IEG, is given only for PMG 425 = fr. iamb. *1 W.²) and to West's ones for the elegiac fragments; from Page are derived, for example, the erroneous γένεο in PMG 357.9 (obviously to be corrected to γενέο, imperative) and ὀρσόλοπος in PMG 393; in PMG 402b, instead of using the minor type (which his own criteria call for), B. adopts Page's textual arrangement, which he defines as ‘misleading’ (p. 654 n. 81); in PMG 417 he uses Page's line-numbering while adopting a different colometry, and in PMG 445 he comments on Page's text (pp. 789–91) while making a more restrictive textual choice; as for the elegiac fragments, in the proposed layout, neither fr. eleg. 3 (that puts into colometry the prosastic structure of Page, PMG 504, or the iambic by Rozokoki [2006], p. 88) nor fr. eleg. 5 (with Θρῃκίης) fit in to the elegiac rhythm; (b) has chosen not to re-order and re-number the fragments; (c) limits himself to the essential notations in the apparatus and collects only the testimonia ‘for Anacreon's life’, ‘but not for the form or content of his poetry’ (p. 93), without critical apparatus.

The ‘General Introduction’ is well structured, thoroughly documented and bibliographically up to date (see pp. 59–91). Well-balanced and convincing are the observations concerning ‘Life’ (pp. 1–11), with substantial confirmation of the traditional chronology and the status of Anacreon as a ‘wandering’ court poet; similarly for ‘Themes’ (pp. 11–16), with the predominance – perhaps due not only to the choices of tradition – of the erotic theme, alongside the divine and purely symposial ones, and their mingling; ‘Models’ (pp. 16–18), with the right emphasis on the role of the Odyssey and Sappho; ‘Performance’ (pp. 18–22); ‘Language and Style’ (pp. 22–8): very opportune, even if apparently incongruous, seems to me the choice of not normalising the language of Anacreon in a more or less Ionic sense, unless in the presence of a clear documentary attestation (cf. PMG 431 μοκλόν but καθεύδει); ‘Metre’ (pp. 29–31), with the obvious exclusion of the elegiacs and iambics κατὰ στίχον, and with the analysis of the iambo-trochaic, aeolο-choriambic (mostly glyconic based) and ionic metres; ‘Reception’ (pp. 31–45), in which very little space is given to the Byzantine, medieval and modern legacy (p. 45); and ‘Fate of the Text’ (pp. 45–54), from pre-Hellenistic times to Alexandrian editions and beyond. The introduction concludes with editorial issues such as ‘Orthography’ (with plausible choices about ‘contractions’), ‘Presentation of the Fragments’ (p. 54) and ‘Excluded fragments’ (pp. 54–6). Questionable is the choice of using the circellus not for dubia, but as a signal that a word used in any form by Anacreon is put into the text in the basic form (cf. PMG 446A–C, 461, 467–74, 477, 479, 480, 482–5, 487); accordingly, I would have prudently done so also for PMG 448, 453–6, 458, 462–4, 466, 481, 486; otherwise, I might have limited myself to pointing out Anacreon's words in the quoting context, as B. does in PMG 444 (where, however, μύρων ἀνάπλεως καὶ γεγανωμένος is unlikely to be Anacreon's sequence: cf. G.M. Leo [2015], pp. 187–8). Not very convenient, also given the ‘Harvard-style’ adopted for the reference system, is the placement of the bibliography (very complete, even though a chronological list of Anacreontic editions is missing), which is preceded by a list of abbreviations, in an internal position, between the introduction and the text and translation.

The chapter ‘Text and Translation’ – preceded by a concise selection of testimonia, with reference to A. Müller 2010 and to the announced complete collection by L. Bucceroni – offers (a) the mere diplomatic transcriptions (except for P. Ryl. 1.35 at p. 157), with apparatuses, of four papyri (P. Oxy. 3695, the very large 3722, 4454 and P. Ryl. 1.35) not included in PMG and SLG; (b) the text (sometimes preceded by a diplomatic transcription) and translation of fragments included in PMG and SLG; (c) the text and translation of the elegiac fragments.

While the notations of the apparatus (suitably updated) are concise, the registration of the witnesses is broader, but most often the relationships of dependence between the witnesses are not indicated (as e.g. those between Athenaeus and Eustathius); also, there is no fragment-by-fragment metrical overview: rather, a metrical section for each fragment can be found in the commentary. The real added value of the work is the translation, which appears consistently after the text (if complete and interpretable) and under the first witness of each fragment (the first two in PMG 446A), sometimes immediately after the Greek, sometimes at the end of the witness series (as e.g. for PMG 348).

After an extensive commentary, the work ends with a ‘General Index', an ‘Index of Greek and Latin Words Discussed’ and an ‘Index of Passages Discussed’: the first is rich and well structured; the second includes combinations of particles such as δὲ δή and phrases such as ἐγὼ δέ, but records verbs sometimes in the infinitive, sometimes in the first person of the indicative present; the third is selective, obviously without Anacreon, but with the Anacreontea; the editors of the texts of the various authors are sometimes indicated and sometimes not. There is no (complete) index verborum and no index fontium, which confirms the exegetical rather than strictly textual framework of the work.

B.'s choice not to include the epigrams among the genuinely Anacreontic texts is correct (cf. p. 35), but perhaps – in these cases – short sections of dubia and spuria could have been included. The case of the Anacreontea (discussed on pp. 36–9) is different; they are rightly excluded from the edition.

The habit of systematically publishing only diplomatic transcriptions of the papyri – except PMG 346 and 347, for which both the diplomatic and the interpretative transcriptions are provided, albeit with some inconsistencies between one and the other, because traces or letters uncertain in the first are given as certain in the second (cf. PMG 347 passim) – and of offering the divisiones verborum and the most reliable integrations only in the apparatus does not always allow one to distinguish at a glance between certain and conjectural elements; also, the apparatuses never give the alternative letters for an uncertain trace (but the commentary most often does), and this does not allow one prima facie to restrict the integrative choices or to evaluate the plausibility of certain under-dotted letters and of certain supplements.

The apparatuses are not always rigorously constructed: for example, in P. Oxy. 3722 fr. 2, the notations at ll. 4f. (Maehler) and 5 (Henry) should be connected, as they are (partially) alternatives for the same portion of text, while in the notation at ll. 8–9 it is necessary to write ‘παρὰ Πυθομά]ν̣δρου̣ Maehler’; in PMG 357 (as in Page) the first, the sixth and the last apparatus notations relate non-homogeneous text segments (see also PMG 367, 371, comm. ad PMG 374 p. 528, 402c.2, 442, SLG 315B, fr. eleg. 2 W. etc.), and the third does not clarify to whom the lesson received in the text (πορφυρῇ) is to be attributed (Fick), as in PMG 361.1 (where ἐγὼ δ’ ἂν οὔτ’ is lectio tradita, but with δ’ by Casaubon preceding Barnes, who accepts, perhaps rightly, the transposition δ’ οὔτ’ ἄν), or in 401.1 (where δηὖτε is by Bergk).

Above all, the habit of indicating supplements without naming the proponent of such a suggestion can be a little perplexing, though some help in attributing the supplements can be found in the commentary. Furthermore, the use of the formula ‘vd. comm.’ (e.g. pp. 118–19, 120, where H. Maehler's supplements are provided only in the commentary) does not necessarily render the apparatuses autonomous, nor does it facilitate their use, especially since the commentary is part of a separate volume. The same goes for notes such as ‘cf. Bernsdorff 2011’ in the commentary, passim.

Conversely, the (attractive) idea that some fragments ought to be considered as part of the same compositions (P. Oxy. 3722 fr. 2.1 and PMG 454; P. Oxy. 3722 fr. 17 c. ii 3 and PMG 437; PMG 347 fr. 1 and 422; PMG 346 fr. 11+3 and 390; PMG 433 and 434) is only advocated for in the commentary (pp. 270–3, 287, 355, 607–8, 767). Much in same way, the fact that in PMG 412 the final question mark was added by Bergk (1834) and that the attribution to Anacreon was proposed by Hermann (1816) is only specified in the commentary (pp. 683 n. 150, 684), without the apparatuses and notes on the fragments retaining trace of it.

The (interesting) supplement ἀσήμω | [ὑπὲρ ἑρμάτων φορέομαι⋅] χειμάζομαι (Maehler) for P. Oxy. 3722 c. i 3f. is proposed (correctly, but s.n.a.) in app. ad PMG 403.2, but not ad P. Oxy. l.c. (see, however, the commentary ad l. on p. 280). Even the semiography is not always consistent and denounces B.'s dependence on the editions of the various testimonia: this is the case with the brackets that indicate deletions, sometimes braces (see pp. 190, 194, 195, 203, 205, 245, 246, 247), sometimes square (see pp. 204, 220).

More in detail. PMG 349.1: Bergk writes δηὖτε Θαλυσίους. PMG 350: there are no compelling reasons, it seems to me, to put ἀνασεσυρμένην into the text (B. also thinks it may be part of a poem in glyconics and pherecrateans on Baubo); better to space both ἀνασύρειν and ἀνασεσυρμένην in the witness (Phot. α 1687 Th.) or have them preceded by the circellus. PMG 351, 353 (and elsewhere): Etymologicum Symeonis is mentioned (following the example of Page) only for cod. V (edited by Gaisford). PMG 355: I would translate ‘An. too uses …’ (κέχρηται … καὶ Ἀν-), and unify ‘Sud’ and ‘Suda’ (better in italics, as title). PMG 356(b).1: μηκέτ’ is Mehlhorn's (1827:107) reading before Fick's (1888). PMG 357: among the witnesses, Hdn. GG III/1 79.13, 159.12 must be added; l.10: the capitalisation of Ἔρωτ’ reflects an idiosyncratic interpretation of the term and of the infinitive δέχεσθαι, in my opinion unnecessary. PMG 377: the presence of δέ does not make immediately necessary the integration of a verb ‘to say’ in the passive (e.g. λέγονται) before ἱπποθόρον δὲ Μυσοί, and the correction of εὑρεῖν in εὗρον (Bergk) would perhaps be more economical than positing for a complex syntax. PMG 379: I would have written παραπετέσθω all in a minor type and a spaced body (not only the ending), since there are many forms of the verb that An. may have used. PMG 388: see also (v. 8) Hesych. θ 1004 L.-Cunn. (not mentioned even in the commentary); for the final lacuna I had proposed (Lirici greci [2011], p. 365) ἐκκαρείς (cf. Ar. Ach. 849–50), ‘perfectly shaved’. Coherent with the effective synthesis of the Anacreontean meters, undoubtedly consistent with the general approach (but certainly not the only possible one), is the interpretation of PMG 388 as composed by variously enlarged anaclastic glyconics, rather than as anaclastic and polyschematic choriambic tetrameters, as it still seems to me preferable, given the accentuated choriambic incipit of the first two lines of each stanza. Fr. 396: ceteris paribus (cf. RFIC 144, 2016, 424), the arguments in favour of μή and against δή for v. 4 (pp. 632–6) all seem controversial and reversible, and the problem remains unsolved (μή, in any case, detracts force from the image of boxing); asyndetic ἔνεικον at v. 3 is not translated. PMG 428: despite the fact that it is attested by Hephaestion and the presence of the incipitary δηὖτε, B. does not affix the beginning-poem sign (so also in SLG 315A, where δηὖτε appears and it is also certain that it is an incipit). PMG 437: the apparatus does not record the (adopted) variant κόκκυξ (Et. Gen. [J], Et. Gud. 333.22–4, not registered) : κόκυξ (Et. Gen. [AB], Et. M.).

Spatial limitations prevent me from engaging in a systematic analysis of the commentary, the best and largest part of the work (exemplary of this is the treatment of PMG 388, on pp. 575–604), even if on occasion the same piece of information is repeated several times, the collection of loci similes may seem excessive, and there might be a need for greater conciseness and focus. In any case, B.'s notes are truly excellent, they do not evade textual problems, they are precise and comprehensive in recognising essential issues (often treated in titled paragraphs, as is almost always the case for the metre, but without a fixed order), scrupulous in the reconstruction of Realien, exhaustive in the word-for-word analysis and refined in the detection of stylistic phenomena (in particular the sound figures): this is undoubtedly a work of high philology.

Only a few marginal remarks: PMG 346: though the commentary is masterful on the whole, the note on fr. 1.13 λεωφ]όρε λεωφόρ’ Ἡρο[τ]ίμη (anaphora can also be significant) seems substantially in dissonance with the ancient exegesis of Suda μ 1470 A. and Eust. Il. 1088.38–9, 1329.34–5 (which document the offensive character of the epithet); fr. 4: as regards the interpretatio metrica (see pp. 330–1), the position of the particles in vv. 1, 7, 8 advises against solution (i) (for which we would have punctuation three times after the first element of the verse) in favour of (ii) or (iii) (in which the punctuation would constitute the central caesura); the use of the aorist form ἐκφυγών (l. 4), in my view, contradicts the idea that the verb refers ‘to a state of liberation from Eros that has not yet been reached’ (p. 337). PMG 357: it does not seem so obvious to me that ‘one of the main oppositions of the poem is that between wild nature and civilization’ (p. 438); rather, the erotic value is indubitable, especially if one interprets the last infinitive – with a more natural syntax and like the ἐπακούειν of l. 8, in turn connected to ἔλθ’, ‘come and/to listen’ – as iussive and as connected (δ’) to γενέο σύμβουλος, ‘advise him and make him accept/to accept’ (‘Be … and admit’, instead, B. p. 173); all of this advises against ‘the identification of Cleobulus and Eros’ (pp. 438–9 and passim) as well as his ‘assimilation’ to Dionysus (p. 449 and passim), and against the idea that the subject of δέχεσθαι is Dionysus and not Cleobulus (p. 452: Meleag. HE 4270, appropriately underlined by B., confirms the traditional interpretation of δέχεσθαι), which also has repercussions on PMG 358, where B. imagines (prudently, but without too much foundation) that the ἄλλη τις (κόμη) of the last verse is that of Eros (p. 456). PMG 395: as for the ‘dying at a young age is second-best etc.’-motif, see also Soph. OC 1224–7; certainly too confident is the statement about Sapph. fr. 58 V. that ‘since the publication of a Cologne Papyrus … the poem is almost complete, and we now know where the poem starts and ends’ (p. 619), because the question of P. Oxy. 1787 fr. 1's additional tetrastic is still open (this also has exegetical consequences for the interpretation of PMG 395, whose similarity with Sappho is a bit overestimated on pp. 619–21). As for the topos of the ‘way of no return’, its origins can be traced back to a near-eastern koine that transcends the boundaries of the Greek world: cf. for example Katabasis of Ishtar (1.5–6: see B.R. Foster, Before the Muses. An Anthology of Akkadian Literature, I–II [1993], pp. 403–9), 2 Sam. 12:23, Job 7:9, 10:20–2, 16:22 etc.

There are more inconsistencies and typos than one might have expected, curiously especially with Greek accents, but they are more than understandable in a work of this size.

Ultimately, Anacreon, the poetic commentator of the luxurious comforts of the courts of the last archaic age, can now finally enjoy a complete, updated, modern commentary, devised by a competent and intelligent philologist, ἀντ᾽ ἐρατῶν δώρων τῶνδε χάριν θέμενος (AP 346.2 = FGE 495).