It has always been hard to resist invoking Phillimore’s quot editores, tot Propertii as the motto of Propertian textual criticism. This new edition of Propertius by Giancarlo Giardina succeeds his edition of 2005 ( Properzio. Elegie. Edizione critica e traduzione. Roma: Edizioni dell’Ateneo ), and is the fourth complete edition of the text to have appeared since the same year.1 If the editor of the recent Oxford Classical Text celebrates an era of ‘diversity and openness in textual choice’, for which he gives us a new motto in quot lectores, tot Propertii,2 Giardina has gone one further as a single reader for whom there is more than one Propertius, and given us a second text which in certain respects makes the new OCT look conservative by comparison. In its preface, the new edition claims to be ‘una ristampa largamente riveduta e corretta’ of its predecessor, which was reviewed for BMCR by Frédéric Nau ( 2006.05.02). That review should be consulted for a description and appraisal of Giardina’s approach to, and vision of Propertius, which in essence are unchanged; the present review will attempt to establish the extent and detail of the differences between the two editions.
The new introduction duplicates roughly half of its content (c. 11 pages) from the edition of 2005 (with minor differences of punctuation and fewer paragraphs).3 The second branch of the manuscript tradition (i.e. A and its descendants) is now called Π (as in Heyworth’s OCT), but Giardina’s position on the history of the text is essentially as it was (no stemma of the tradition is postulated in either edition). In particular, the new introduction reaffirms (2010: 17-19) its predecessor’s disbelief (2005: 14-17) in the third branch of the manuscript tradition hypothesised independently by Butrica (as Χ) and Heyworth (as Λ), but now focuses on the arguments advanced by the latter in the intervening period, restating (this time without acknowledgement) Goold’s observation that the manuscripts said to descend from the hypothesised Χ/Λ always fail to provide solutions where the undisputed branches Ν and Α/Π are corrupt.4 Giardina goes on to take issue with Heyworth’s dismissal of error duplication between the manuscript T, hypothetically descended from the disputed Λ, and the two undisputed branches. Advocates of the third branch will respond that Heyworth provides many more examples where Τ shows its independence of errors in the undisputed Ν and Α/Π, including a sample of ten from the two surviving quires of A (spanning Prop. 1.1.1 to 2.1.63).5 T conversely duplicates an error found in A only at 1.18.19 where, in giving ardor for Ν’s arbor, it commits a single-letter slip of the kind which in this edition Giardina thinks palaeographically plausible (see further below and, e.g., p. 34: ‘[n]essuna delle mie letture congetturali urta drammaticamente contro la verosimiglianza paleografica’). Differences of opinion aside, Giardina is not necessarily more likely to reject readings found in the manuscripts of the hypothetical Χ/Λ family, which he terms recentiores, and in which he finds many credible readings.
The introduction also revisits the (mis)use of loci similes by (unnamed) textual critics (2005: 20-21; 2010: 22-3): the earlier discussion of 4.5.64 is replaced by 1.3.16 ( osculaque admota sumere et arma manu) which, for Giardina, constitutes a locus dissimilis with Ov., her. 13.141 ( arma dabit, dumque arma dabit, simul oscula sumet), the latter not suggesting the innuendo of the former (some may disagree), the former not being as explicit as it should, and therefore to be desanitised as gaudiaque admoto sumere rara sinu. That such is the implication of the Propertian passage many readers will agree, but some may regret an emendation which defuses the latent violence that gives the poem its edge.6 Whereas the earlier edition highlighted such wholesale corruption of entire words (2005: 25), the newer edition also lays emphasis on letter-for-letter corruption (2010: 26-34), providing to this end a series of examples (2.4.13, 2.9.17-18, 3.5.3-4, 2.13a.1, 2.26b.23-4, 3.4.1, 4.5.21, 4.7.95-6: the first three of these have been incorporated in the new OCT) where sense can be restored without stretching the limits of palaeographical credibility (‘uniscono alla corretezza sul piano semantico una certa solidità sul piano paleografico’ [2010: 26]).
Having reaffirmed in this way the editorial principles behind the earlier text, Giardina reiterates that his edition is ‘fortemente, spesso audacemente, innovativa’ (2010: 34). This is even more true of Propertius 2010 than it was of Propertius 2005. The sheer number of departures from the earlier text may account for why these are nowhere listed in the later edition. Some 654 differences, to be posted separately on the BMCR blog, represent a departure from the 2005 text on average once every six lines approximately (this figure rises to once every 4.5 lines approximately for Book 1, and drops to once every ten lines approximately for Book 3). Of these differences, 377 (58%) represent new emendations proposed by Giardina; in 174 cases (27% of the differences), the new edition withdraws emendations proposed by Giardina in 2005. The codicum consensus is reinstated 87 times (13% of the changes), and abandoned 355 times (accounting for 54% of the changes).
It may be imagined that those who would emend the text only with the gravest of hesitation will be troubled that so many new readings should be introduced, in just five years, to a text that had already been radically edited. Such readers will at least welcome the restitution of the textus receptus at, e.g., 2.3a.22 or 2.18b.10, where adiunctos sedula lauit equos alludes once more to Callimachus hy. 5.9-10, or again at 3.1.30, where Parim is the lectio difficilior to Giardina’s former parem. Conversely, the loneliness and isolation that are integral to the Waldeinsamkeit would dissuade most from emending the universally transmitted inflections of solus at 1.18.4 and 2.19.7. Those pursuing metapoetic readings are not likely to agree that the textus receptus should be mistrusted at 3.3.15 (where the substitution of flumine with munere denies this programmatic elegy an important Callimachean metaphor), or at 4.9.50 (where it is intelligible literally and significant metaliterarily that Hercules’ transvestite hands are indeed ‘hard’), or again at 4.10.3 (where the idea of Propertius’ generic ascent is lost if we read intendo rather than ascendo); at 2.1.40, we do not need to be told that Callimachus does not sing inflato pectore in order to infer that he always sings, as the textus receptus tells us, angusto pectore. It seems unlikely that the adjective turpis is not the correct reading in at least one of the six places where it is rejected in this edition (1.16.22, 2.3a.4, 2.16.36, 4.4.1, 4.5.4, 4.7.55), especially since three of them constitute a possible thematic nexus in Book 4. If loci similes do have weight (the apparatus in this edition provides a service in swelling their ranks) but are to be treated with caution (as the introduction exhorts), then we might not find more compelling the parallels Giardina cites for, e.g., his fulgeat usque at 4.2.46 than those cited by Gregory Hutchinson ad loc. for the transmitted langueat ante.7 At 2.12.13 harundo is backed up with Virgilian and Ovidian parallels, but the loss of imago perhaps takes with it a Lucretian resonance (cf. DRN 4.1052-62). The parallels given for feruet sanguis may tempt some to replace the challenging uernat sanguis at 4.5.59, but they will lose their foretaste of the vernal imagery in the famous lines which follow. Nor need intertextual connection depend exclusively on lexical duplication: lentus (Francius) need not be adopted at 4.8.75 for the phrase cultus in umbra still to recall Ecl. 1.4, as Ovid ( Ars. 1.67) and Martial (11.47.3) testify. The justification for emending 1.17.4 on the basis of similar cases of dactylic confusion in Propertius that are each already emended by Giardina is not without circularity.
Response to individual interventions is of course a subjective matter, and will depend inter alia on how risk-averse a given reader is. There is perhaps profit to be made from betting on, e.g., candidus at 1.20.32 (if a dolor! is not a Gallan ejaculation) or cluerant at 2.13b.38 (rarer and therefore possibly more likely than fuerant). At 2.32.55 femina quae (for dic mihi quis) looks like a winner with quae dea at the start of the following line. The parallel with Dido at 4.4.69 suggests that divine agency need not necessarily be involved here (although it may invert the Virgilian moment(s) that it is a goddess, rather than Giardina’s uirgo, who feeds the flame of culpability). Those who invest in the transposition of 4.8.19-20 may stand to double their margins by adopting Giardina’s tum for the transmitted cum.
Then there are the bigger risks, the bolder interventions, e.g., Cimmerias for Memnonias (1.6.4), fugis militis officia for uago fluminis hospitio (1.20.10), moecha for uerba (2.5.28). At 4.5.60, attractive parallels entice the investor to consider nam uolucris labitur usque dies for ne quid cras libet ab ore dies. Yet the real risk may, at times, be to keep faith with the textus receptus, challenging though it is. Paradoxically, many of Giardina’s adventures will produce a less adventurous Propertius, at the cost (some may feel) of what is most characteristic of the poet (see also, e.g., fulgens. . .templum for fugiens. . .portus at 4.6.15, nimis. . .improba for tamen…ianua at 4.11.85). To this extent, Giardina 2010 will make us think again about where, despite the warnings, we take the Propertian text for granted, and work harder to justify our decision not to emend, when that is our choice. However, caueat lector : the decision not to print cruces desperationis (see p. 34), together with the sheer frequency of intervention, make it unlikely that any reader could, through this edition, become familiar with Propertius as known to most.
In editing Propertius so radically, Giardina can enjoy the reflections of G. P. Goold: ‘Most men are more concerned to be slightly wrong, if they are to be wrong, than to be audaciously right, if they are to be right: there are perils in being right with a Galileo, whereas one risks nothing by being wrong in a conformist world of Ptolemaic error.’8 This new edition is indeed ‘largamente riveduta’, then, but amid so much revision it has given itself difficulty following through on the preface’s second claim: as if to prove the inevitability of scribal slips and the consequent necessity for emendation, a considerable number of errors has crept into the new text and apparatus (and these too must necessarily be recorded).9
Notes
1. Other editions since 2005, in chronological order, are: Viarre, S. (2005) Properce. Elégies. Collection Budé. Paris: Les Belles Lettres 2005 (an edition judged to be conservative by Giardina 2010: 27); Heyworth, S.J. (2007) Sexti Properti Elegi. Oxford (omitted from Giardina’s list of editions on pp. 37-8). The same period has also seen publication of a further three texts and commentaries on individual books: Fedeli, P. (2005) Properzio, Elegie Libro II. Cambridge: Francis Cairns; Hutchinson, G.O. (2006) Propertius. Elegies Book IV. Cambridge; Heyworth, S.J. and Morwood, J.H.W. (2011) A Commentary on Propertius Book 3. Oxford.
2. Heyworth ( op. cit., n. 1), lxv.
3. The only typographical error of potential consequence in the introduction occurs on p. 15 (four lines up) where ‘APLPZ’ should read ‘AFLPZ’. Other errors are inconsequential: p. 18 (line 19): ‘essere’ not ‘esssere’; p. 28 (line 6, in the quotation): ‘its’ not ‘is’; p. 32 (n. 1): ‘Cambridge’ not ‘Cambrideg’.
4. Goold, G.P. (1990) Propertius, Elegies. Cambridge Mass. 13
5. See Heyworth ( op. cit., n. 1), xxx.
6. See, e.g., the reading of Propertius 1.3 in Greene, E. (1998) The Erotics of Domination: Male Desire and the Mistress in Latin Love Poetry. Norman: University of Oklahoma Press. 51-9.
7. See Hutchinson ( op. cit., n. 1), ad loc.
8. Goold, G.P. (1967) ‘ Noctes Propertianae‘, HSCP 71: 59-106, at 92.
9. Errors detected in the text and apparatus can be found in the comments of this review on the BMCR blog.