Bryn Mawr Classical Review

Bryn Mawr Classical Review 1998.12.05

Stefano Martinelli Tempesta, La tradizione testuale del Liside di Platone. Pubblicazioni della Facoltà di Lettere e Filosofia dell' Università degli Studi di Milano 173. Sezione di Filologia Classica 6.   Florence:  La Nuova Italia Editrice, 1997.  Pp. 340.  ISBN 88-221-2905-9.  L.56,000.  

Reviewed by David Murphy, The Nightingale-Bamford School
Word count: 2231 words

Anyone who wants an introduction to the textual tradition of Plato should read this erudite and far-reaching study. Martinelli Tempesta digests several decades of "dialogue-by-dialogue" examination of the Platonic text in this investigation of all the witnesses of the Lysis. Along the way, he shows in fascinating detail the mind and habits of scholars who worked on Plato from late antiquity to the High Renaissance. The author shares valuable results of work in situ on numerous, lesser-known witnesses, and his archival work brings to light the provenance of a number of them. M.T. reads everything, he weighs evidence carefully, and he sets out clearly the reasoning behind his conclusions. I subscribe to almost all of them.

M.T. pursues three principal aims. The first is to classify all the witnesses to the Lysis as far as Stephanus: manuscripts, humanist Latin translations, printed Greek editions, and the indirect tradition. His second goal is to mine these witnesses for information about the scholars who produced them. The third is to deduce how the three families of our medieval MS. tradition are related. M.T.'s conclusion on this score, perhaps the most controversial of the book, is that the three primary MSS. represent different minuscule transcriptions of a text that served as the standard in late antiquity, however much that late antique text had undergone contamination.

The primary MSS. of the Lysis turn out to be the familiar trio BTW, i.e. Oxon. Bodl. Clarke 39 (=B), A.D. 895, Ven. Marc. append. cl. IV.1 (coll. 542) (=T), mid-tenth century, and Vind. suppl. gr. 7 (=W), eleventh century. The tenth-century Vat. Pal. gr. 173 (=P), which offers only excerpts of this dialogue, is related to W but independent. The apographa generally maintain the relationships to their exemplars and to each other that they display in other dialogues of the first seven tetralogies. M.T.'s arguments for their stemmatic relations are sound, and he makes good use of material evidence.

As is the case in many dialogues, the contemporary hand B2, whom M.T. takes to be Arethas, imports into B numerous readings that we find also in W. M.T. relied on the photofacsimile to distinguish corrections made by the first scribe from those made by B2 (pp. 9-11). The correction at 216b4 looks to me like the work of B2 not B, and against what is said on p. 10 n. 27, I have found while working on B in situ that B2 makes many corrections per rasuram, while the first scribe on the other hand will introduce a correction above the line without erasure. From B-B2 was copied Vat. gr. 226, from which in turn descend Vat. Urb. gr. 32 and Erlangen 1227 A 4. M.T. incorporates the results of his earlier demonstration that Erl was used by Regiomontanus (Johann Müller of Königsberg) to learn Greek in the house of Bessarion (Martinelli Tempesta 1995(1)), although he qualifies his former belief that Leonardo Bruni was the copyist of Urb.

Only two MSS. descend from W in the Lysis: Prague Lobc. Roudnice VI Fa 1 and Vat. gr. 1029. Having shown that Lobc is a copy of W and that Vat. 1029 is a copy of Lobc (Martinelli Tempesta 1992), M.T. now marshalls the evidence against the recent theory that Lobc is an imitation of the hand of W and not the production of W's scribe (pp. 129-41). In 1992 he reported that two lacunae in W are supplied by a later hand, and examination of microfilm now leads me to think that this hand is W3, active around 1300 (cf. Murphy 1995, 155-62): 209c1 ἐν οἷς ... κωλύουσι] om. W, add. W3im; 212e5 οὐκ ἔμοιγε δοκεῖ, δ' ὅς] om. W, add. W3im.

All the other MSS. are copies of T. M.T.'s classification of the T apographa in the Lysis largely replicates the results that students of other dialogues have reached. The bulk of T apographa descend through Par. gr. 1808 (=Par, thirteenth cent.; M.T. hesitates to accept the earlier date proposed by Brockmann 1992). One of the most influential descendants of Par in other dialogues, late thirteenth-century Scor. y.I.13, has no progency in the Lysis (pp. 95-97). A second copy of T is Par. Cois. gr. 155 of the fourteenth century, parent of yet two more MSS. I now think M.T. is right to deny that Cois is a direct copy of T (p. 41). Vat. Urb. gr. 80 and part of Ven. Marc. gr. 186 descend from T through a lost, third intermediary, while from a fourth independent copy of T, also now lost, M.T. derives five MSS. M.T. has already brought one of the latter to our attention, viz. the little-known Wroclaw Bibl. Univ. Akc. 1949/60 of the early fifteenth century (Martinelli Tempesta 1995(2)). M.T. identified its scribe as Bartolomeo Aragazzi da Montepulciano, a conclusion that Christian Förstel had also reached independently (Förstel 1994). M.T. provides now a description of its twin, Lond. B.L. Harley 5547, copied by Zomino (Sozomeno) da Pistoia. Copied within the circle of students of Manuel Chrysoloras and Cencio de' Rustici, these two MSS. provide evidence of how Plato was being studied in N. Italy and Switzerland just after 1400.

The Lysis was translated into Latin by Pier Candido Decembrio, Marsilio Ficino, Pier Vettori (M.T. promises a critical text), Janus Cornarius, Jean de Serres for the Stephanus edition, and an anonymous Jesuit, whose translation is bound into a copy of the printed Greek edition of Vettori. Making use of two dedicatory MSS. of Ficino's translation, M.T. shows how certain of Ficino's readings correspond to corrections introduced by Bessarion or John Sgyropoulos in Cois and Marc. gr. 186. M.T. demonstrates that we cannot yet pinpoint what MSS. Ficino may have used beyond Flor. Laur. 85.9 and Conv. soppr. 180; Ficino's working exemplar was neither of those (pp. 157-76). M.T. is right to note that Musurus' Aldine Greek text does not mark any significant advance over the material that came to him from Par. gr. 1811 and Marc. gr. 186. A preponderance of errors that the Aldina shares with Par. 1811 in Chrm. and Smp. makes it look as though Musurus relied on that codex more than on Marc. 186, and I am not convinced by M.T.'s attempt to set these errors aside. He may however be right that Musurus used not these MSS. but some lost recension already made from them (pp. 183-84).

The relation between the MS. families is one of the vexed problems of Platonic textual studies, for the tradition is heavily contaminated, and much material has been lost. Although M.T. directly addresses the Lysis and, to a lesser extent, the Symposium, his discussion ranges over many dialogues. In chapter 6 he presents evidence against the views of Gerard Boter and Christian Brockmann that TW descend from the lost first volume of Par. gr. 1807 (=A), a ninth-century minuscule MS. (cf. bibl. pp. 213-16, 262-64.). In chapter 7, M.T. goes on to hypothesize that TW do not share a common hyparchetype; instead, BTW descend in a tripartite stemma from a single source, and mutual contamination infects these three branches. He is right to contend that maiuscule errors shared by BTW show that our medieval MS. tradition has a common source (pp. 266-68), and B's many unique maiuscule errors and errors of word division in many dialogues do confirm M.T.'s claim for its stemmatic independence. The hard part is to establish that T and W go back independently to the same source.

M.T. first argues that BTW represent three distinct minuscule transcriptions. He appeals to 1) unique maiuscule errors in the single witnesses B, T and W, and 2) absence of minuscule errors shared by any two of the three. M.T. appears to be right about 2), and this fact, in combination with the almost total absence of significant TW minuscule errors in other dialogues, weakens Boter's and Brockmann's theory. Not all errors cited in support of 1) from the Gorgias, Lysis and Symposium (pp. 216, 226 and 265) are equally significant, for M.T. recognizes that errors of word division or accentuation can arise from minuscule script (p. 216 n. 19, 265 n. 53). I would take this cautionary note further, given the fairly common recurrence of maiuscule letter forms in minuscule script by the mid tenth century. I cannot accept M.T.'s two maiuscule errors of T, for the faulty word division at 219d5 can have occurred during transcription of a minuscule MS., and 222b4 (not b8) is in fact the correct reading (the hand that added this reading in the margin of W looks to me like the first scribe not W2, pace M.T.). Moreover, I find one compelling maiuscule error in W, viz. 216d1 διαδύεται BT: διαλύεται W (perhaps also 208e4). Still, some of M.T.'s single maiuscule errors of T and W in other dialogues retain their weight (Grg. 501e10, Smp. 212c7, 223b9 of T; Grg. 521c4 of W), and M.T. is right to point out that shared maiuscule errors need not have come into T and W from a minuscule exemplar. I have found a few maiuscule errors of W in Chrm., most notably 156a5 ἀκριβοῖς] ἀκριβοῖο and 169d9 που] μου. T and W may indeed represent distinct minuscule transliterations.

Next, M.T. examines the TW errors of Lysis and invokes coincidence or contamination to deny that we must posit a TW hyparchetype to explain them. I am more impressed than he is by 216e3 τοῦ τοιούτου B: οὐ τοῦ τοιούτου TW and 218b6 πάνυ γε ἐφάτην· νῦν ἄρα B: νῦν πάνυ γε ἐφάτην ἄρα TW (ἐφάτην· ἆρα W). No one has shown that T cannot have transmitted one member of certain pairs of variants, of which W transmits both. On the other hand, there are precious few conjunctive errors from which to prove that TW share a common intermediary, especially in view of BW errors like the apparent misreadings of maiuscules at 204a3 ἂν T: δὴ BW and 222b4 τι ὡς TWim: πως BW (cf. above). It seems to me that one can make the data in Lysis consistent with bipartition or tripartition.

It is not so clear however that TW do not share a maiuscule hyparchetype in e.g. tetr. iv or Phdr., where TW agreement in error is massive. M.T. approves the attempt by Carlini 1972 to renounce the bipartite stemma that he had previously defined for tetr. iv. Carlini instead extended tripartition from Phdo., where it appears valid, to tetr. iv et al. by appealing to contamination to explain TW errors. M.T. does not really refute W.S.M. Nicoll's criticism of this degree of reliance on contamination as an explanation of errors, however right it is to remind us to consider the effects of horizontal transmission of variants. Most workers on tetr. iii-vi have favored a bipartite stemma (besides those mentioned, cf. Moreschini and the OCT editors), while the support of Dodds and Bluck for tripartition was quite tentative (cf. bibl. on pp. 262-64). I do not in fact share M.T.'s degree of skepticism about bipartition in dialogues like Alc. I or Phdr., where TW transmit essentially one tradition, and I wonder how far M.T. thinks we may go in invoking contamination as an explanatory principle. The practical import of this question for the editor is not great, however, for in an open tradition like this, Maasian stemmatics offer little guidance.

Finally, M.T. is an accurate collator. He collated the entire dialogue by autopsy in the Greek Vaticani, Marciani (except 189), Laurentiani and Neapolitani (except II.C.32) and from photographs of the others. I have checked most of his reports of B and W against the photofacsimile and microfilm, respectively, and have found only occasional misreports of accents or division of words. One of these causes doubt about one of M.T.'s maiuscule errors, for at 207c2 ἀμφισβητοῦμεν] ἀμφισβητοῦ· μεν re vera W (M.T. reported ἀμφισβητοῦ· μὲν, pp. 138 and 226), it is not so clear without the grave accent that W is guilty of faulty word division. Conflicting reports are given of W at 208d4 on p. 133 (correctly, εἴης ἢ)) and p. 226 (ᾗς ἢ), wrongly, where, as also on p. 15, M.T. with Burnet misses B's word division; it is actually ῃση). At 208b8, not only B but also W omits T's δὲ; at 219d2, B's μάλιστα ἄλλα finds an unreported counterpart in W's μάλιστα ἀλλὰ (p. 15).

But these are quibbles. This book has vastly increased my knowledge, not only of the Platonic textual tradition, but of the tradition of its study. It has been a constant companion during my own work on the Platonic text. One looks forward to many future contributions in what promises to be a signal career.


Brockmann, Christian. 1992. Die handschriftliche Überlieferung von Platons Symposion. Serta Graeca 2. Wiesbaden: Dr. Ludwig Reichert Verlag.

Carlini, Antonio. 1972. Studi sulla tradizione antica e medievale del Fedone. Rome: Edizioni dell' Ateneo.

Förstel, Christian. 1994. "Bartolomeo Aragazzi et Manuel Chrysoloras: le codex Vratislav. AKC. 1949 KN.60." Scriptorium 48:111-21 (pl. 8-11).

Martinelli Tempesta, Stefano. 1992. "Collazione del Liside." In Studi su papiri e codici filosofici. Platone, Aristotele, Ierocle, 83-86. Studi 129. Florence: Accademia Toscana di scienze e lettere "La Colombaria."

-----------------. 1995 (1). "Un codice platonico usato per apprendere il greco." Studi Umanistici Piceni 15:127-44.

-----------------. 1995 (2). "Un nuovo codice di Bartolomeo da Montepulciano: Wroc. Ms. Akc. 1949/60." ACME 48:17-45.

Murphy, David J. 1995. "Contribution to the History of Some Manuscripts of Plato." RFIC 123:155-68.

Read Latest
Index for 1998
Change Greek Display
Books Available for Review

HTML generated at 13:26:01, Friday, 03 April 2009