Since Mynors’s edition of the Panegyrici Latini almost fifty years ago, studies on the text of Pliny’s Panegyricus still depend largely on his results. There may be new insights particularly into the history of the text, problems still to be solved, and even disagreement in detail, but on the whole his work remains fundamental. 1 This new edition by Rosario Morena Soldevila offers something more, inasmuch as it is furnished with a long introduction and further material for a better understanding of the text, but also something less, because as an edition it falls back in many respects to a pre-Mynors level. It is divided into four sections: a long introduction of 104 pages, a list of sigla, the Latin text together with an apparatus criticus, an apparatus of sources and loci similes and a Spanish translation en face with notes, and a short index of proper names.
The introduction which is to give "some keys" ("algunas claves", p. XIII) is an informative read. It is strong when it comes to Pliny’s biography and his works, even his lost works, because Moreno Soldevila makes skilfully use of his letters as an autobiographical source. She sheds light on a number of topics crucial to our interpretation of the text, esp. the images of Trajan, Domitian and Nerva, the role of libertas or the importance of religious aspects including the question of the person of Trajan in relation to the gods. It is shorter when she deals with the genesis of the speech and its literary models, and it is perhaps too short as regards the historical setting of the Panegyricus, which fills only one page and remains, without any indication of ancient sources, rather abstract. In the chapter on language and style (I.8) some interesting observations on aspects of colloquial Latin and on poetic elements can be found as well as an analysis of panegyric formulas, but one is surprised to see that prose rhythm and in particular clausulae have received no more than four lines of the author’s attention in which the most recent literature mentioned in the text is the dissertation of W. Baehrens (1910), and two cases of clausulae are classified incorrectly.2 Remarkable is the extensive bibliography: in fourteen pages it covers the topics discussed and, although completeness is hardly possible, it may well serve both as a first step into the problems related to the Panegyricus as well as an instrument for further research.3 The edition of the text itself is prepared by short chapters on Nachleben, the transmission of the text and its editions with special interest in Spanish authors, manuscripts and translations, as well as a "Valoración" of two pages, i.e. views on the Panegyricus from Alfieri to 20th century scholarship, and a summary of the single chapters of the speech.
For the constitutio textus the manuscript tradition, the most important editions and the "bibliografía" of textual criticism have been reviewed according to the author (p. LXXXVI). It is not made clear whether the manuscripts have been collated directly, but in the end this turns out to be a point of minor consequence. The real problems of this edition begin with stemmatic issues. It is one of Mynors’s achievements to have seen that the Upsaliensis C 917 (A), before him regarded as an important independent witness, is a descendant of the Harleianus through the Napocensis bibl. univ. lat. 7, which afterwards was demonstrated in detail by Lassandro – and now it is back again, filling an apparatus criticus with its own Sonderfehler like an inconstructible tu instead of te in 22,3 or a lacuna in 68,1, giving the impression that there is a problem in the transmission where there is not (e.g. A’s eorum in 14,5 or at in 27,5), and quite often simply agreeing with its ascendant H. The Upsaliensis is an eliminandus to be taken into account for conjectures, of which only one has found its way into our text (remotioris 39,5) or if no ascendant were readable.4 The same applies to the descripti of X, e.g. when we learn that the Vaticanus lat. 1775 (W) omits in spem aliorum in 27,1. There is not much consequence when on p. CXV of Moreno Soldevila’s edition the Upsaliensis is listed as H’s descendant or W as descendant of X: after H follows A with a preceeding unde, after X unde determines X1, W’s family. As to the question of where the good readings we find in the editio Cuspiniani come from, Moreno Soldevila (p. LXXXII) adopts Mynors’s view that they derive from an independent, but lost manuscript, without mentioning Mynors or indicating that there is a question at all. But if we accept that any reading of the editio Cuspiniani can derive from this lost manuscript independent of X or H, then we must know about the variae lectiones of the Cuspinianaea at every point. In fact, what we see in the apparatus criticus is that often readings of the editio Cuspinani are left out, for instance montes in 50,1 or in ita 81,3, as if they were proven to be conjectures which could be included by the editor whenever she thinks them to be interesting. The expression Cuspinianus which is used throughout the book does not make the picture clearer because it normally denotes "conjecture by". In the perusal of conjectures, however, a similar inconsequence becomes visible. Livineius is the scholar to whom the honour seems to be given of first conjecture in the edition: "caperetur Livineius" (1,1, instead of capere). Yet in his edition (Antverp 1599) it is simply what he has printed in his text, whereas in the margin he refers to the reading capere he had also found, and others long before him had printed caperetur.5 In 2,6 his splendid conjecture is presented as "principum Livineius" – correctly, but in the same manner as the indication in 1,1. In 4,1 quo is followed by "W A.Baehrens W.Baehrens Kukula", but the three scholars mentioned simply adopted a reading W has (more likely a correction by w, Tommaso Parentucelli 6) into their text, a reading which can be found in Livineius’s text, too, who is not mentioned. The confusion is certainly not diminuished by a practice to be found here as well as throughout the rest of the book, i.e. by creating chains of scholars’ names who adopt a conjecture into their text which gives the wrong impression that a particular conjecture has been made by several scholars independently; besides, often the criterion for inclusion in these chains is not clear. So Kukula (Leipzig 1908 and 1912) whose editions are not comparable to those of A. Baehrens (Leipzig 1874) or W. Baehrens (Leipzig 1911) appears disproportionally often. As a result it becomes hardly distinguishable what is a genuine conjecture and what is a mere indication of a "probat", and the indications are not as reliable as they could be. The reliability of the edition is furthermore challenged by the fact that at 4,7, 45,5 and 59,4 it gives no indication of a problem in the manuscript tradition although we have festinatis against festinantis; possumus against possimus; at against ad; and according to Mynors also in 29,3: nequiquam in M, corrected to neqicquam by the Itali. Of new conjectures there are two: in 23,3 Moreno Soldevila suggests reading senatu instead of senatus parallel to equestris ordinis flore, not impossible as a panegyrical exaggeration; in 62,3 she inserts non before prosit.
Examples of the issues discussed could be multiplied, but the multiplication would not make the point clearer. Despite useful remarks in the introduction, the apparatus locorum similium and the notes to the translation7 which help to interpret the text, a reader of the edition does not feel on safe ground. The introduction can stand for itself, but the edition is better not used without Mynors’s or Lassandro’s at hand.
1. Mynors’s edition, Oxford 1964, was praised by M. Winterbottom as "model edition" in his summary of the Panegyrici Latini in L.D. Reynolds (ed.), Texts and Transmission. A Survey of the Latin Classics, Oxford 1983, 289, a kind of Festschrift for Mynors himself. D. Lassandro in his edition of the Panegyrici Latini, Turin 1992, follows him closely (see esp. XXVIf.), but V. Paladini-P. Fedeli, Panegyrici Latini, Rome 1976, p. XXVIIf. (who do not give a text of Pliny) regard the Harleianus 2480 (H) as independent of the reconstructed Moguntinus. A problem neither solved nor even fully analysed is e.g. the exact stemmatic position of the editio Cuspiniani, Vienna 1513, as some of its readings seem to go back to a tradition not otherwise known, see Mynors p. viii; Lasasandro XXVII; Paladini-Fedeli p. XXXIIff.
2. See p. LXVI. Humanitate detrahitur (4,7) and det sed accipiat (87,4) are not examples of the double cretic but of a resolution in a cretic-trochaic sequence.
3. Some titles, however, are missing like e.g. G. Seelentag, Taten und Tugenden Traians. Herrschaftsdarstellung im Principat, Stuttgart 2004, a book that is hard to ignore even if an author does not always agree with it. On the other hand, the bibliography is rich in references to Spanish literature.
4. See Mynors p. vii and D. Lassandro, I manoscritti H N A nella tradizione dei Panegyrici Latini, BPEC 15 (1967) 55–97, an article which is not mentioned by Moreno Soldevila.
5. Schwarzius has a note on p. 1 of his edition (Nuremberg 1746) about them.
6. See Manfredi, Un’editio umanistica dei Panegyrici Latini Minores: il codice Vaticano lat. 1775 (W) e il suo correttore (w), in L. Belloni-G. Milanese-A. Porro (eds.), Studia classica Iohanni Tarditi oblata, Milan 1995, 1313–1325, who is nowhere referred to in the book. There are other cases in which w seems to be misrepresented as W: see the apparatus on 17,1; 24,5; 86,6.
7. The reviewer regrets that he is not competent to evaluate the details of the Spanish translation.