We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 11 February 2009
Over the last fifty years the world of the palaeographer has been revolutionized by the widespread use of photography. Today a scholar can study a microfilm of almost any codex in the western world in the comfort of his home and compare it with any number of other codices within a matter of minutes. It is no longer necessary to travel long distances, set aside large blocks of time, and spend substantial sums of money in the collation of manuscripts. This fact should encourage modern palaeographers to review the work of their predecessors who were denied these blessings to see if the work of the past lives up to today's standards.
1 Cerocchi, P., ‘Prolegomena ad Xenophontis Hipparchicum’, SIFC 6 (1898), 471–92.Google Scholar
2 See, for example, Marchant, E. C., Xenophontis opera omnia, Tomus V: Opuscula (Oxford, 1920)Google Scholar, Praefatio ad Hipparchicum, ‘Ad textum Hipparchici… recensendum fundamenta optime iecerunt tres viri Itali, Cerocchi, Tommasini, Pierleoni, quorum nomina semper ab omnibus Xenophontis operum minorum studiosis celebrabuntur.’ Since many of the manuscripts of Hipparchicus also contain De re equestri, V. Tommasini's work in SIFC 10 (1902), 95–119Google Scholar can also be helpful to the student of the former opusculum. But because his findings are virtually the same as Cerocchi's, Tommasini's collations and stemma also deserve reevaluation. See also, Pierleoni, G., Xenophontis opuscula (Rome, 1933).Google Scholar
3 See Cerocchi, , p. 474Google Scholar n. 1: ‘ceterorum notitiam compluribus humanis atque liberalibus viris debeo… qui in usum meum hos codices ad Hipparchicum carptim contulerunt.’ It is in the area of collation and in the relationships of manuscripts based on these collations that Cerocchi's article is weak. In other respects his work is still of great value.
4 See Cerocchi, , p. 474.Google Scholar My B is his b, considered by him to be a fourteenth-century codex. See my article, ‘The Mysterious Manuscript A of Xenophon's Cynegeticus’, Hermes 117 (1989) 157–66Google Scholar, where I offer a new evaluation of the relationship between B and A (= Vindob. phil. gr. 37).
5 My placement of B in the overall stemma agrees with that of Cerocchi, p. 492.
6 For descriptions of these codices see Cerocchi, , pp. 480ff.Google Scholar He knew very little about L and could therefore not give it its rightful place in the group. Both L and H are featured as representatives of the second family of Hiero manuscripts by Haltinner, D. and Schmoll, E. A., ‘The Older Manuscripts of Xenophon's Hiero’, Revue d'histoire des textes 10 (1980), 234–5.Google Scholar Codices O and W do not contain Hiero, so we have here a better picture of the whole group and its relationship to the γ group which follows.
7 Cerocchi, , pp. 482ff.Google Scholar, graphically demonstrates that lacunae in H are represented in anp. Because it lacks evidence of these lacunae, he sees s as a likely gemellus of H. I deal with this matter in my text. Cerocchi's list of H-group errors (p. 484) is correct at 1.3, 6, 21; 3.5; 4.15; 5.1, 12 and 8.18. At 1.4 and 7.4 he lists as H readings items which belong in the λ-group list. The reading at 1.16 is found also in W, that at 5.13 also in OW, that at 1.19 also in OW and, in the same place, L omits the reading in a lacuna. Where H has κα⋯ at 3.2, OL read ⋯π⋯ and W ⋯ν. I have not been able to see the Munich manuscript, but its trouble with H lacunae makes it a certain member of the group.
8 I owe my information about Hiero to Leverenz, Lynn, ‘The descendants of Laur. 80.13 in Xenophon's Hiero’, SIFC 7 (1989), 12–23.Google Scholar The information about Cyn. comes from an article by Schmoll, E. A., ‘The fragmentary manuscripts of Xenophon's Cynegeticus’, Syllecta Classica 1 (1989), 21–5.CrossRefGoogle Scholar He informs me that the fragmentary opening lines of Cyn. in the Vienna codex are a duplicate of the fragment carried in H. W has a slightly longer fragment (see below). See note 4 above for the derivation of a in Cyn.
9 Again, Cerocchi's list of r (his d) readings (p. 481) is correct at 3.2, 9; 4.8, 10, 14; 8.5, 25 and 9.5. The supposed error at 7.3, the first time the phrase occurs there, is spurious. Also, at 1.19 the omission in L is marked by a lacuna, but there is none in r. The second reading at 3.2 is also found in O. The addition of μ⋯ν at 4.12 is a λ-group error. The second reading at 9.5 is also found in P of the γ group. Cerocchi also gives a list of readings which he believes prevent our seeing r as a copy of L. The reading at 1.25 seems to affirm rather than deny the relationship. At 8.23 also the variant seems to originate in L where the parent has a strange abbreviation for -
ν of ⋯ρμν. The variants at 4.3 and 7.13 are itacisms whose corrections are not beyond the talent of a fifteenth-century scribe. At 4.1 the scribe has simply repeated the compound of νο⋯ω used a few lines above and he has introduced errors at 5.8 and 11. At 4.8 Cerocchi got unreliable information from Gardthausen about L which has ἥξοντες. At 9.7 the λ group reads συνελθ⋯ντων. The scribe of r saw that it was not ‘if the gods convene’ that was needed, but ‘if the gods will’, and changed to συνελθ⋯ντων, close to the BM γ reading συνελθ⋯ντων We must also admit that some contamination may be involved in the readings of this fifteenth-century witness. At 5.13 the whole λ group has ἔχον and the return of r to ἔχειν of BM γ is suspicious. Also, at 1.7, while r has εἰκ⋯ς ⋯στιν in the text (a BM γ reading), it also has εἰκ⋯ς ε
(the λ-group reading) in the margin.
10 This stemma differs from Cerocchi's (p. 492) in having HLOW derive from a single parent. The λ-group readings assure this derivation. Codex r (his d) is surely derived from L, if not a direct copy of L. B.M. 5110 is a derivative of H. I have not attempted to ascertain whether anps are independent derivatives of H or whether one serves as the source for one or more of the rest.
11 Diller, A., ‘The Greek Codices of Palla Strozzi and Guarino Guarini’, Journal of the Warburg and Courtauld Institute 24 (1980), 320.Google Scholar
12 Deuling, J. K. and Cirignano, J., ‘A Reappraisal of the Later ABS Family of the Tradition of Xenophon's Hiero’, to appear in Scriptorium in 1990.Google Scholar
13 The designation γ for this group comes from an unpublished dissertation by Schmoll, E. A., Xenophon's De Venations: a Collation, Stemma and Critical Text (The University of Iowa, 1982).Google Scholar For a more accessible description of these manuscripts see Serra, G., ‘La tradizione manoscritta della Costituzione degli Ateniesi dello pseudo-Senofonte’, Atti e Memorie dell’ Accademia Patavina 91 (1978–1979), 82–3.Google Scholar See also Cerocchi, , pp. 486ff.Google Scholar Serra has done a masterly job of accounting for earlier scholarship on Ath. Because the λ group lacks Ath. and because he does not know about the SM relationship (see below), Serra's stemma is not accurate.
14 Cerocchi cites a few readings (p. 487) which occur early in the opusculum to establish this group. Most are insignificant and poorly documented. At 1.1 P also has ἄρξαας. At 1.2 C has δ⋯, not δ᾽. At 1.4 P agrees with the rest of the group and with BM against λ. At 1.5 what amounts to a π reading is in agreement with B. Again at 1.7 the γ group agrees with BM against λ. At 1.8 C does not omit τε, but P does. At 1.10 we have another instance of a λ-group error.
15 On p. 488 Cerocchi gives some readings which indicate to him that P cannot be a copy of F (his g). He is correct, but both have ἄρξαας at 1.1. P and Fa have πορρηθ
ναι at 1.14, both have συμβουλε
σαι at 1.18, παραλε⋯ψω and ⋯παλ⋯ττοιτο at 3.3. P evidently introduced κα⋯ at 9.5, because no contemporaneous or older manuscript has it. The upshot is that neither F nor P is the source for the other. In his dissertation (n. 13 above), lacking separative F readings in Cyn., Schmoll considers P a copy of F. This is no longer tenable.
16 On pp. 487–8 Cerocchi shows that, although q (his V) is very close to F (his g), q cannot be a copy of F. The closeness comes from π readings passed on to F and P. The actual source of q is P.
17 See Haltinner, and Schmoll, (n. 6 above), p. 235.Google Scholar They did not collate P, but I have looked into the citations of Hiero which they think point to contamination in F and find that they occur also in P. This is further proof that it was the source of FP which was contaminated. Since Hipparchicus does not occur in the manuscripts of Hiero designated ABSC, it looks as though M or one of its offspring is the source of the contamination.
18 Cerocchi's E group (see p. 492) corresponds to my γ group. He sees all five extant manuscripts as brothers, an idea I hope has been dispelled here. Cerocchi (p. 488) also sees little value for text reconstruction in this group. Not only is this not true, but much can be learned about the head of the whole second family of opuscula by noting the order of contents in γ.
19 The best description of M is Serra's (p. 81), but see also Cerocchi, (pp. 474f.)Google Scholar who was very disappointed in the quality of M's text. The identification of Bessarion's hand in the codex is mine.
20 In an unpublished dissertation, Xenophon's Agesilaus: a Collation, Stemma and Critical Text (University of Iowa, 1975)Google Scholar, Rosemary Wieczorek says that M derives from S through a lost intermediary (pp. 38ff.). Deuling and Cirignano see no need for such an intermediary in Hiero. I am quite sure, in fact, that the variant on κεκτσθαι at 11.15 in Hiero in S was written by the scribe of M himself.
21 For how S is related to A see my article, ‘Correction and Contamination in Xenophon's Hiero’, SIFC, Terza serie, 6 (1988), 68–76.Google Scholar
22 This theory gets no support from A. G. Roos who, in his 1967 Teubner edition of Arrian, places Marc. 511 in the second family of manuscripts and Ambros. E 11 inf. in the third (pp. xix and xxii).
23 See Cerocchi, , pp. 475ff.Google Scholar, Serra, , pp. 81f.Google Scholar, Haltinner, and Schmoll, , p. 234Google Scholar, Deuling, and Cirignano, , Schmoll dissertation, p. 25Google Scholar, Wieczorek dissertation, pp. 38ff. Wieczorek presents substantial evidence to show that Marc. 368 does not follow M in Ages. See also my unpublished dissertation, The Manuscript Tradition of Xenophon's Hellenica (Indiana University, 1967), pp. 56ff.Google Scholar, for the idiosyncracies of this manuscript in Hellenica.
24 This is, in effect, where Cerocchi puts M. His E = my γ and his B my β. Serra (p. 84) does very little with the abbreviated Ath. manuscripts, but he too places M close to γ (= his m). Schmoll in his dissertation takes both M and γ independently back to the archetype (I maintain his Φ for the archetype), but he had not yet evaluated the abbreviated Cyn. manuscripts which fall into the γ group. Today he feels that M and γ are more closely related than he had earlier supposed. I should also state here that β and γ could be the same manuscript in different states of repair. M could have been copied from γ before it began to fall apart and δV after.
25 Let the reader beware of other renditions of these readings in the various editions of Hipparchicus.
26 The β errors are the same as the γ-group errors listed above, except for the first and last which M lacks.
27 These are the γ-group errors listed above.
28 See the good readings of B alone listed at the beginning of this article.
29 Serra, pp. 110ff., attempted this same exercise, but because he tried to accommodate a list of the opuscula found in Diogenes Laertius and because he did not know about the S influence upon M, his ingenious shot in the dark falls short of the mark.
30 In the following list H2 refers to the older portion of Laur. 80.13 and M2 to the contents of M which do not derive from S.
31 Because M and the γ group are free of errors characteristic of the λ group, lost codex λ cannot be codex β in a state of deterioration.
32 Hiero is also contained in the companion volume. The scribe of C therefore omitted it as well.
