Three New Homerica on Papyrus - Renner

8/2/2019 Three New Homerica on Papyrus - Renner

1/28

Department of the Classics, Harvard University

Three New Homerica on PapyrusAuthor(s): Timothy RennerReviewed work(s):Source: Harvard Studies in Classical Philology, Vol. 83 (1979), pp. 311-337Published by: Department of the Classics, Harvard UniversityStable URL: http://www.jstor.org/stable/311104 .

Accessed: 02/05/2012 04:00

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

Department of the Classics, Harvard University is collaborating with JSTOR to digitize, preserve and extend

access toHarvard Studies in Classical Philology.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=dchuhttp://www.jstor.org/stable/311104?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/311104?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=dchu


2/28

THREE NEW HOMERICA ON PAPYRUSTIMOTHY RENNER

OF all Greek authors whose works are represented among thepapyri, none has survived in a greater profusion and variety oftexts than the poet Homer. Pack's catalogue of literary papyri and thesupplementary inventory by F. Uebel together list 516 published papyriof the Iliad and 158 of the Odyssey, which comprise fragments of booksof all grades and dates from the early Ptolemaic through the Byzantineperiod.1 Corresponding to, although lesser in bulk than, the enormousnumber of papyri of the two epics is an extensive and variegated bodyof Homerica, or subliterary texts pertaining to Homer. These consistof the glossaries, paraphrases, lexica, commentaries, anthologies, andother aids that either were meant to assist the teacher, student, or readerof Homer as study and reference tools or, in the case of some paraphrasesor summaries, were composed by the student as part of an exercise orassignment. Three varying specimens of such Homerica from thepapyrus collection of the University of Michigan are here published forthe first time.2A glossary may be conveniently defined as a list of words or lemmataarranged in the order of their occurrence in the original text and eachbearing a gloss which contains one or more literal or nearly literal inter-pretations of the expression in question. A particular class of Homericglossary is that known as scholia minora, in which the glosses sometimescontain more than one explanation of each lemma but do not as a ruleattempt an elaborate or learned discussion after the manner of commen-taries or scholia proper. Texts of this type, of which several dozen ex-amples on papyrus are in existence, are also united as a group by detailedsimilarities in their methods of explanation and in the contents of their

1 R. A. Pack, The Greek and Latin Literary Texts from Greco-Roman Egypt,2nd ed. (Ann Arbor 1965) 49 ff, 157; F. Uebel, Archiv fiir Papyrusforschung21(1971) 174 ff; 24/25 (1976) 191 ff.2 The papyri were first studied as part of my dissertation, Literary Papyri in

the University of Michigan Collection (Ann Arbor, Univ. Microfilms 1974).Thanks are thus due to O. M. Pearl, the chairman of my doctoral committee,and H. C. Youtie, as well as to the others who have since then made helpfulsuggestions.


3/28

312 Timothy Rennerglosses. These show that the scholia minora are the remains of an amor-phous, fluctuating, and mainly anonymous body of straightforwardlexicographical material that was popular rather than scholarly innature. The first papyrus in our series is a glossary of the scholia minoratype written in the first or second century A.D. The date makes it oneof the earlier examples of texts from this body of scholia, the interpreta-tions of which from at least as early as the first century A.D. began tofind their way into a variety of other scholia and lexicographical worksincluding Apollonius Sophista and, eventually, the Byzantine compila-tion known as the D-Scholia.3The second papyrus, excavated at Karanis, is a fragment of a second-century A.D. lexicon to Homer arranged alphabetically by the first twoor three letters of its lemmata. Homeric lexica of this type, which showmany affinities with the interpretations of the scholia minora and of theD-Scholia, have been found only occasionally among the papyri. Acomparison with other lexica preserved in manuscript suggests that theMichigan fragment, like five other lexica on papyrus, represents anearly copy of the Lexicon Homericumof Apollonius Sophista. The workof Apollonius has been transmitted to us in a single manuscript and inshortened form, but much of the material in it was utilized by Hesy-chius and by other lexicographers in late antiquity.The third papyrus, from an anthology or summary of Iliad E-T,also comes from Karanis and gives some more indication of the kindof literary study that took place in that town. The piece is interestingboth because its writer quotes Homeric passages often and because itdates from the Ptolemaic period.4

3 For the nature and interrelationships of the scholia minora on papyrus, theD-Scholia, Apollonius Sophista, and other lexicographical compilations contain-ing Homeric material, see A. Henrichs, "Scholia Minora zu Homer I," ZPE 7(1971) 97-149, especially the introductory material, 99 ff. Earlier importantstudies are those of Wilamowitz, Hermes 23 (I888) 142-147 and A. Calderini,Aegyptus 2 (1921) 303-326.

4 In the line notes, lexicographical parallels have been cited where possibleaccording to the abbreviations used by Henrichs, ZPE 7 (1971) 117 ff. ForEustathius the edition of Stallbaum (Leipzig 1825-29) has been followed exceptin the case of Iliad A-I, where M. van der Valk's first two volumes (Leiden 1971-76) may now be consulted. Of the scholia to Homer other than the D-Scholia,those to the Iliad are cited after H. Erbse, Schol. gr. in Hom. II. (Berlin 1969-77)and those to the Odyssey after Dindorf. The references to Photius follow theedition of Porson (Leipzig 1823). "Orion" refers to the Etymologicum of Orion(ed. F. Sturz, Leipzig 1820). Square brackets around a reference indicate thatthe lemma in question appears there in a different inflected form from that inthe papyrus.


4/28

Three New Homerica on Papyrus 313I. Scholia minora To Iliad A 1-9P. Mich. inv. 1588; 17.7 x 15 cm; late first or secondcenturyA.D.;provenanceunknown

Purchased in 1924, this papyrus contains two columns of lemmataand explanations from a glossary of the scholia minoratype on the open-ing verses of the Iliad. The text, written parallel to the fibers, hassuffered from surface wear. Column I is essentially complete, butColumn II has lost portions from the right side of all its glosses. Onthe back are two to three columns, badly marred by wear, of a gram-matical text. The writing, in a well-formed hand of the type often seenin carefully executed documentary texts, may date from either the firstor the second century, although the latter seems somewhat more likely.5The arrangement of lemmata and glosses in parallel columns that areseparated, except where the gloss is a lengthy one, by blank spaces ofseveral letter-widths, follows the format of the majority of other similartexts. When a gloss requires more than one line, the writer begins thesecond line below the lemma but indents slightly. With the exceptionof some items the meanings of which should have been obvious to eventhe most inexperienced reader of Homer, the great majority of the wordsfrom the verses covered by the text have been glossed. Four publishedpapyri containing scholia minora for all or part of these lines of the Iliadare available for comparison:i. P. Berol. inv. 5014 = Pack2 1158 (fifth century A.D., Panopolis?),ed. U. Wilcken, Sitzungsb. d. k6n. pr. Ak. d. Wiss. zu Berlin, Phil.-hist.Kl. (1887), 818 f; cf. Wilamowitz, Hermes23 (1888) 142 ff. Covers A I-,2.62. P. Achmim2 = Pack21159 (third or fourth century, Panopolis), ed.U. Wilcken, ibid. 816 if; also discussed by Wilamowitz (cf. above). CoversA 1-21 (summary of A precedes).5 Many of the letters have cursive forms, although the writer is not alwaysconsistent with regard to a given letter. More often than not, each letter in aline is written separately, but there are a number of cursively linked pairs orgroups of letters throughout the text. The writer's use of his space was reason-ably generous (upper margin 1.7-2.3 cm, lower 1.7 cm - both perhaps origi-nally deeper). Similar hands are Schubart, P. Gr. Berol. Papyrus 22 (A.D.135 ?),Papyrus 24 (A.D. 148), Papyrus 28a (second century A.D.),P. Lond. I1z8a (Facs.Vol. I, P1. 52; A.D. 145), 140 (Facs. Vol. I, P1. 21; A.D.69-79). The grammaticaltext on the back is crudely written and difficult to date palaeographically butwould be compatible with the period in general.6 On this and the following papyrus cf. also A. Ludwich, "Ober die Papyrus-Commentare zu den homerischen Gedichten," in Index Lectionum I (K6nigs-berg 1902).


5/28

314 Timothy Renner3. P. Oslo II 12 = Pack2II6o (second century, Theadelphia ?). CoversA 5-24.4. P. Oxy. XLIV 3207 (third century, Oxyrhynchus), ed. pr. F. CarterPhilips, BASP 8 (1971) 91 ff. Covers A 4-18.

With the exception of 2, the above glossaries display a fullness of cover-age of the Homeric words that is comparable, or nearly so, to that of theMichigan papyrus. It is to be expected that a higher proportion of wordsshould be glossed in scholia minora pertaining to the very beginning ofthe poem than in those covering a passage further along, where thereader would have built up a greater familiarity with the epic diction.This is in general confirmed by a survey of the fifty or so publishedpapyri of scholia minora to both the Iliad and the Odyssey.7In the commentary, I have cited exhaustively the corresponding ex-planations offered by the four papyri listed above, all of which usuallyshow a close affinity with the Michigan text. Also included are refer-ences to other lexicographical sources, among which the D-Scholia,Apollonius Sophista, and Hesychius generally contain the closest paral-lels to the glosses of the papyri. It has become clear that the scholiaminora are essential to the textual, historical, and comparative study of7 Of this total group of scholia minora texts for both Homeric epics, fullyfifteen, or about 30 percent, pertain to Iliad A - a testimony to the frequencywith which that book was read.Twenty-seven papyri in Pack2 which contain scholia minora are enumerated

by Henrichs, ZPE 7 (1971) 104, n.24. Pack2 1176, which contains a glossary toA 364-371 as well as a prose paraphrase of part of the book, should probably beadded to the group; likewise Pack2 II70, although unpublished, probably con-tains scholia minora. In the same note Henrichs also lists six Cologne andMichigan texts first edited by him in "Scholia Minora zu Homer" (ZPE 7[1971] 97 ff, 229 ff, 8 [1971] I ff, 12 [1973] 17 ff) as well as nine more papyriwhich appeared after Pack2. Of the latter group, P. Oxy. ined. with glossary toA 4-18 was published by F. C. Philips in BASP 8 (1971) 91 ff and is now P.Oxy. XLIV 3207; and P. Oxy. ined. with glossary to A 302-323 was edited byS. Stephens as P. Oxy. XLV 3237. Further additions to the list of scholia minorapapyri must now be made as follows: P. Oxy. XLV 3238 (to A 405 ?-538, B385-393; early third century A.D.); P. Hamb. inv. 736 verso, ed. Th. Vlacho-dimitris, ZPE 1 (1973) 65 ff (to B 61-222; second century); P. Oxy. XLIV3158 (to E 655-725, 782-822; second or third century); XLIV 3159 (to H 4-8o?;preceded by hypothesis; third century); a wooden tablet in Alexandria, ed.H. Riad and J. Schwartz, Chr. d'Eg. 43 (1968) 114 ff (to A 3 1-48; preceded andfollowed by passages from the Iliad; date uncertain); P. Vindob. gr. inv. 39940,ed. P. J. Sijpesteijn and K. A. Worp, ZPE 15 (1974) 153 ff (to 0 320-412,52o-633; late second century); P. Oxy. XLIV 3160 = P. Strassb. inv. 1401,ed. J. Schwartz, ZPE 4 (1969) 175 ff (cf. W. Luppe, ZPE 27 [1977] IOI ff; toa 441-444, ) 2-300, 426-434; includes hypotheses; third century). Also, P. Oxy.XXX 2525 verso according to Lobel's brief description is a scholia minora text(to B 201-218; second century).


6/28

Three New Homerica on Papyrus 315these and other lexica and scholia. This can be seen most clearly in thecase of the D-Scholia, which are the end product of many centuries ofanonymous and very simple explanatory efforts such as we find in thepapyri. Finally, for comparative purposes four prose paraphrases ofthe Iliad, compiled by Byzantine scholars but heavily influenced by thescholia minora tradition, are cited after the partial edition by A.Ludwich, Aristarchs homerische Textkritik II (Leipzig 1885) 490 ff.Intended as translations the function of which was to assist in thecomprehension of Homer by the reader, these are:Par. P The Psellos Paraphrase (also known as the Bekker Paraphrase),which of the four is closest to the D-ScholiaPar. M The paraphrase of Manuel MoschopoulosPar. G The paraphrasewritten by Theodoros of Gaza, who essentiallyreworked the Moschopoulos paraphrasePar. A The interlinear paraphrasecontained in Codex Venetus Graec.822

The Michigan papyrus contains no breathings, accents, or otherlectional aids. Iota-adscript is omitted. The writer makes use of a sus-pension once, in I.ii. Discernible errors of spelling as such appearonly in 1.2 and 3 and are minor ones. In 1.4 someone, perhaps a secondhand or pen (M2) using a more watery ink than that of the main text,corrected an accusative pair to the genitive so as to bring the gloss moreinto line with his notion of clarity. It appears that M2 may also havebeen responsible for at least the gloss of a short entry inserted betweenlines I.2 and 3 as well as for a supralinear addition to the gloss of line 3.The heading giving the book number above column I may have comefrom the same source.8

COLUMN1.

. jvov XOAOV77V 8v.ov (A I)2 .EL.S aSEo.~}&{} " VHVEL (I)2a qfq Mov aa V V(I)o3 H-r1,ArtaSKw> H-)EWS VLtovEcYEL (I)

SEo[?v]JXLAAE~aj8 With the heading we may compare the similar notation marking the transi-tion from one book to another in P. Colon. inv. 2381 (ed. A. Henrichs, ZPE 8[1971] 3 ff) and in P. Berol. inv. 11634 (ed. G. Poethke, Forsch. u. Ber. d. Staatl.Mus. z. Berlin 8 [1967] 105 ff). P. Colon. inv. 2281 (ZPE 7 [1971] 229 ff) has asubscription of an analogous character.


7/28

3 6 TimothyRenner5 ovAoeLLvv oAEGpLav (2)pwpta 7oAAa (2)

Axatos Tro0L AA7wTv (2).q.f,:.toq ,


8/28

ThreeNew Homericaon Papyrus 317Top margin: at edge of papyrus and above to, a small loop as from aletter level with ].[ in marginof ColumnI; belowand to left, partof avertical (stray?) 2. .: perhaps a letter lost afterwards 9. Inmargin o left, faint traces orminga largeoval(strayor a marginal ign?)A~rpwF ap. i. Stog: o has cross-stroke like cancellation; letter aboves may be E (SVt8os ?) I5. . [: high dot

Col. I i. opy'v P. Achmim2.20, Par. PG; 3py-qv, dAovE'7T1ovovD. See also Par. A, Phot. s.v., Ba. 301.6, Ep. Hom. AO 282.16, Ap.Soph. 112.24, Ep. Hornm.P 296.29, Schol. A ad 7 67. For Ovptdvs anexplanationwe may compareSchol. AT ad A i drEp 7TEaVK7SrPCOTOVMdv ri-v oAvvOos,t-r 0"/A776~KOV?So *x, oVTo 7TrpwrTOVpy, OvtO&sXd'AosKdo-rStqv!S. Our papyrus may have had another word addedabovethe line.2. Closest s Ap. Soph. 10.9,which hasJ8e, V'jtL, 8v'vcuLv tepe7rocl1cULaTre&ace.See alsoHsch. a 1253,Par.PMG, D., Schol.bT. The papy-rus may have had supralinearwriting.o{s}was perhapswrittenunderthe influenceof eafollowing;see Mayser/Schmoll,GrammatikI.12.i83;F. Gignac,A Grammar f the GreekPapyriof theRomanandByzantinePeriods Milan 1975) I 125 f.2a.0.pMovra:

So P. Achmim2.20, P. Berol, inv. 5014 recto I, Par.A, Hsch. 0 156, [Schol. A]. Mov^a.AE'yESrEqv KaAAKrT-qv.; cf.Par. M.7(Tpwvy.c.: Probably 7TrapwVvvuK--,,eferringto the lemma inline 3. This adverb is attested only in the works of Irenaeus,a nearcontemporaryof our papyrus, and in scholia; see Stephanus, Thes.Ling. Gr. and Lampe, PatristicGreekLexicons.v.3-4. ro vo 70-roH7qA,'wsPar. PMGA; 7r rat&S701oHbAE0'ws.Berol.inv. 5014recto2; tIooH~TAl'w78's D. Foro written in place ofcwas in our papyrussee Mayser/SchmollI.12.75 f; GignacI 276 f.5. So P. Achmim2.20, P. Berol. inv. 5014 recto 4, D., Par. PGA.Cf. Hsch. o 1758, Ap. Soph. 124.12, Schol. bT, Ep. Hornm.O 316.6,Ep. Hornm.P 300.26-301.5. oAwdv-qiv- AEOpa[vP. Freib. Ic.i2.6. So P. Berol. inv. 5014 recto 6, Ap. Soph. 114.12, Hsch. t 1i89o,Par. PG, Ep. Hom. AO 270o.17and 284.1. 7) tvPpl' 7-S o7TAAa.Achmim2.21 (cf. D.). See also Ba. 304.25, Phot. s.v., Ep. Hornm.P302.32.7. So P. Berol. inv. 5014 recto 7, D., Par. PMGA, Hsch. a 8815.7TaPrESL r' "IAtov'Tpa1ndE'avres p. Soph. 50.4.8. So P. Berol.inv. 5014recto8, Par. P. KaKa AV7Tas .; &Ay?tara"KaaUTp~'pEL8El s7- KaKC Ap. Soph.21.18. Cf. Hsch. a 2796, Ba.65.19, BekkerAn. Gr. I 375.1, EM 58.io.


9/28

318 Timothy Renner9. So P. Berol. inv. 5014 recto 9, P. Achmim 2.21, Hsch. E 670, Ba.

207.25. TrpoKa7l-rE-KEV,-rot EToloaEv p. Soph. 63.14.Cf. EM

319.30,Apion 233.9.10. So P. Berol. inv. 5014 recto io, Par. P. rrcqaodhhAovsD.Ii. Probably larXvpobV'Xovs4x(aui^v).'rXvpo dbXovs . Achmim 2.22,Ap. Soph. 93.18; cf. Hsch. t I 16, Apion 242.16, Ep. Hornm.P 303.19.IorXvp&s. Berol. inv. 5014 recto II (12 has OvXs'-- s' vuX&s),ar. G,[Ba. 264.30], Phot. s.v.; cf. D., Par. A. See also Par. PM, Ep. Hom. AO207.25.12. So D., Ap. Soph. 14-3, Hsch. ca1755, Par. PMGA.13. E'fEXaAEVs not attested elsewhere as an explanation. 7po8E'OcE0pEv.E/dbowEL S'E'r77v ,per' d3&vr' CdivCv'a7TdACELaAp.Soph.I35-3I; cf. Hsch.7r 3485, Ap. Soph. 89.28. Contrast D., Par. PMGA, Ep. Hornm.O369-4, Schol. bT.14. Between -.pwwv and . /.'prE there would be room for about sixletters. But the absence of any traces of writing in this area suggeststhat the writer omitted part of the gloss. Perhapsthe exemplarhadA'r7EE E62v Lr7E&vOpda'wovr the like. ir-,v iLpOE'wv&v3p v P. Berol. inv.5014 recto 13, D., Par. PGA. 'jpws~ 077LEOS. vvao'r.. yEVvatoS Phot.s.v., Ba. 252.19.

,jpows- vva-,rds.TaXvpos.Ewvatos.

arEpdsHsch. ij 871.

7poE~. of &'abEpoVIES&pE.ollLOEOt. Iv3pES-,yEVaOL Hsch. - 867.wpooar...' yTvoS r63VvvOpdrrrov Schol. A. See also Ap. Soph. 84-32.15. So P. Berol. inv. 5014 recto 14, P. Achmim 2.22, D., Par. PM(and, with minorvariations,GA).16. JAKu'atza, rrap&ytzlraP. Berol. inv. 5014 recto 16, P. Achmim2.23, D., Ba. 217.5. P. Oxy. XLIV 3207 Front I may have had[IU7ap~,yta'a,] JAKdcRa-tzra.ther sources each include one or bothof these equivalents: Hsch. E2251, Ap. Soph. 67.7, Par. PMGA, EM332-57-

EAA.ara: For the remainsof the two uncertainletters,see figure I.The firstresemblese; the second could be a (the diagonalstrokebelowdoes not go well with the t and could be from the downward-curvingtop of a, although I cannot find other examples of a made in thatfashionin this text). A wordEAEzLais not attested.However,the hapaxUAEtta s given in Schol. A ad P 667 as an equivalent for oAwp. erhapsthe writer or his exemplar ntended to write JAE'/.ra but insertedan a(for similarmistakessee GignacI 131).17. So P. Berol. inv. 5014 recto 16, D., Par. P. 'r0olt7UEPar. M.KaCEUrKEVaE Par. GA, Hsch. 7 707; cf. Ba. 386.2.i8. With minor variations, so P. Berol. inv. 5014 recto 17, D., Par.PMGA, Hsch. K 4578.


10/28

Three New Homerica on Papyrus 319

~j j~(1


11/28

320 Timothy Renner5. So P. Berol. inv. 5014 recto 22, P. Oxy. XLIV 3207 Front 5, D.,Par. P, probably P. Oslo II I2 I.5. See also Par. GA, Schol. A (gloss,Dindorf).6.

]To.TE.7T &v& XP. Oslo II I2 1.6. 'viv&pv P. Berol. inv. 5014recto 23, Par. PMG. rrp]6C-rov. Oxy. XLIV 3207 Front 6. Cf. D.,Hsch. 7 169. rTTpdoTEpovp 70Tovvac rrdTOAE1Lovar. A.7. So Par. PM, Schol. bTt. B]'a&r7[auavP. Berol. inv. 5014 recto 24.S&Eariqaav, v''K^OsHsch. 8 1342, P. Oxy. XLIV 3207 Front 7, probablyP. Oslo II 12 I.7. eSt'ar av, E'XwplrOaavHsch. I1335. See also D.,Par. GA.8. Perhaps

'p..cev.[ES?, AovwKacVres,as in P. Achmim 2.24.

qLAOvEKacav-rESD., Par. PMGA (0., 8vFKs P. Oslo II I2 1.8).9-10. Perhaps Arp(


12/28

Three New Homerica on Papyrus 321Soph. 117.32, Hsch. ~ 128, Par. PG. av]vE/OaZAEv. Oxy. XLIV 3207Front 3. E'/aAEcAE,rvviqE .caveAEtVTOlr7UEvar.M.E'rol&)aEv

SAdovELKcaacar. A.16. So P. Oslo II 12 I.15, P. Oxy. XLIV 3207 Front 14. The spaceprecludes a lengthier gloss such as 0 7'rsig-oro Kat AC'TraS' iLrAdMJwvD. (cf. Par. PMGA).17. The traces are difficult to read meaningfully, but the gloss musthave been o0iOs ydp as in P. Berol. inv. 5014 verso 6, P. Oxy. XLIV3207 Front 15 (cf. P. Oslo II 12 I.I6), D., Par. PMGA.

2. HOMERIC LEXICON (APOLLONIUS SOPHISTA)P. Mich. inv. 545ia; Fr. I: 5 x 5.5 cm; Fr. 2: 6.7 x 13 cm;secondcenturyA.D.; Karanis

Obtained from the excavations at Karanis during the 1928-29 sea-son,9 these two fragments of a lexicon contain respectively three andnine entries, many of them nearly complete, from a part of the workdealing with Homeric words in vroAv-,7rop-,and probably 7roa-. Thetext is written parallel to the fibers in a practiced book hand whichprobably belongs to the second century, although the latter part of thefirst could not be absolutely ruled out.10 Except for a few apparentlystray bits of ink, the back of the papyrus is blank. The correspondence,in whole or in part, between several of the entries and those of the solesurviving medieval manuscript of the Homeric lexicographer ApolloniusSophista, the tenth-century Codex Coislinianus 345 (referred to in thecommentary as Ap. Soph.), suggests that the Michigan text, likeseveral other papyri, comes from an earlier and in some respects fullerversion of the lexicon of that author. The papyrus would then representa copy of the work made within a few generations of the lifetime of9 The excavation number is 28/C65*E/A. On literary papyri from Karanis, seen.2o below.10The letter forms are generally similar to those of P. Oxy. XVII 2079, P.Oxy. 2317, and Seider, Paldographie der griechischen Papyri II, Pap. 31 (allassigned to the second century). But these hands are more formal, upright, andregular and less cramped horizontally than that of the Michigan papyrus. Thewriter of the latter at times crowds and reduces the size of his letters as heapproaches the right edge of the column, but he may not have succeeded inkeeping this edge an even one. However, the left edge appears to have beenrelatively straight. The surviving intercolumnar space of 1.5 cm and lowermargin of the same depth (perhaps originally deeper) in Fr. 2 suggest an averageformat for a literary text. The column width would have averaged 5.5 cm.


13/28

322 Timothy RennerApollonius, who was probably active in the second half of the firstcentury A.D. at Alexandria."If the lemma of gloss 12 of our papyrus was rrdaot,as suggestedbelow, it would appear that this lexicon was arranged according to thefirst three letters of each lemma. However, we should beware of assum-ing that this principle was strictly adhered to throughout; for whileeven a few completely alphabetized lexica are known to have existedduring this period, ordering according to the first two letters appearsto have been the commonest system.12 Much later, the Codex Coislini-anus itself follows strict rules of alphabetization only for the first twoletters, although groups of lemmata which are ordered on a three-letterbasis or even in absolute alphabetical order do occur.13Five other papy-rus fragments of Homeric lexica dating from the first through the fifthor sixth centuries A.D. have been plausibly attributed to Apollonius'work. Most of these are alphabetized by two or three letters; one, thelatest in date, is fully alphabetized.14

11 The version of the lexicon in the Codex Coislinianus has been cited afterthe edition of I. Bekker, Apollonii Sophistae Lexicon Homericum (Berlin 1833;repr. Hildesheim 1967). For general information about Apollonius, see thearticles of Cohn "Apollonios " 8o), RE II I 135 f and Tolkiehn, " Lexicographie,"par. 20, RE XII 2 2445. There is further discussion, with references to recentstudies, by A. Henrichs and W. MUiller in Collectanea Papyrologica: StudiesPublished in Honor of H. C. Youtie I, ed. A. Hanson (PTA Bd. 19; Bonn 1976),27 ff.

12 On the alphabetical ordering of lexica in the Roman and early Byzantineperiods, see L. W. Daly, Contributions to a History of Alphabetization in Anti-quity and the Middle Ages (Brussels 1967) 27 ff; J. J. Keaney, "Alphabetizationin Harpocration's Lexicon," GRBS 14 (1973) 415 ff. Papyrus fragments ofancient lexica are discussed by M. Naoumides, "The Fragments of GreekLexicography in the Papyri," Classical Studies Presented to B. E. Perry (Ill.Univ. Sts. in Lang. and Lit., Vol. 58; Urbana 1969) 181 ff; on alphabetization,see 187 ff.13 Lemmata in rroA-and rrop- comprise two relatively extensive series in thecodex, where the rroA- roup of ten lemmata (of which all but two are in woAv-)is interrupted on three occasions by entries in rrot- rwov-, or rrorr-, but the rrop-group of six lemmata is not interrupted at all. The following summary of thelemmata in Ap. Soph. 132.31-134.9 will serve to illustrate (lemmata whichappear in the Michigan text or can be restored there with probability are under-lined): 7TroAVKAKrpOL,ToAvTOc/OVOS,r7TOALOvE a(i'qpov, rroAvKAr0Zat,7ToAtVKEa~7TO,7rOLI'7V,roAv/L7-r77S,TroAVK/ro70V,r6VOS,TOAVLVE,7TOAVapVw,TOVTo7TOpOLr,T07TOL,7rooAAz, 7TOpEV,7'OpK'qgs,oplsr, rropcVp,, opSCa 'rLso,opcf'p'EOs

O vc7rog,7TroU , rrWvta.14 These texts are: (i) Bodl. Ms. gr. class. e. 44(P) = Pack I1217 (first orsecond century A.D., provenance unknown; nine entries in E X-, X-, and ta-;two-letter alphabetization), ed. E. W. B. Nicholson, CR ii (1897) 390 ff, cor-rected by M. Naoumides, TAPA 93 (1962) 243 ff. (2) P. Oxy. XXX 2517


14/28

ThreeNew Homericaon Papyrus 323Of the twelve lemmata in the Michigan papyrus, eight (glosses I, 3,4, 5, 6, 7, 8, 12) occur also in the Codex Coislinianus, provided weaccept the probable restorations in glosses 4, 8, and 12 and allow forthe difference of inflected forms in glosses 5 and 12. Of the eight, onlyone (gloss 7) has an explanation absolutely identical in wording to thatfound in the codex. The explanation of another (gloss 6) is much shorterthan in the codex but summarizes an interpretation of Apion which isa major component of the gloss there. Glosses I, 3, and 5 each preservean additional synonym or comment which has been lost in the codex;but in each case the codex has a somewhat expanded version (includingone case of a quotation illustrating Homer's usage) of the commentswhich it shares with the papyrus. Explanations which differ from theircounterparts in the codex appear in gloss 8 and gloss 4 (as far as thelatter entry is preserved). The remains of gloss I2 demonstrate that ithad a discussion closely related to the brief and presumably abridgedexplanation of the lemma given in the codex. Also noteworthy are corre-spondences of the ordering of the lemmata between papyrus and codex:glosses i and 3 appear adjacent and in the same order in the codex,and the series formed by glosses 4, 5, 6, and 7 also appears there a bitlater on. Finally, the four lemmata (glosses 2, 9, Io, and i i) which donot occur in the codex are nevertheless included in Hesychius, wheretheir explanations bear a strong resemblance at least in part to thosegiven in our papyrus. Thus while we have seen that some of the ex-planations of the papyrus are abridged compared with those in thecodex, in terms of the number of lemmata in these sections the papyrusrepresents a somewhat fuller version of Apollonius. We can alreadyconclude from the other papyri attributed to Apollonius that a varietyof such editions, richer in lemmata than the codex, were in circulationbetween the author's lifetime and the incorporation of a version of his

work into the original compilation of Hesychius in the sixth century.(second century, Oxyrhynchus; fourteen entries in OvL-,00-, Op-,probably twomore in Oay-; two-letter alphabetization), attributed to Apollonius by K. Alpers,Hermes 94 (1966) 430of. (3) P. Oxy. XLIV 32o6 (second century, Oxyrhynchus;34 entries in tLa-,8a8-, LE-, 8aL-, 8at7-, 8K-, VLY-, tLO-; three-letter alphabetiza-tion), ed. pr. J. W. Shumaker, BASP 7 (197o) 59 ff. (4) P. Cairo inv. 502o8 =Pack2 1218 (third century, Oxyrhynchus; nine entries in Ev-;two-letter alpha-betization), ed. W. G. Waddell, Mem. Inst. Fr. d'Arch. Or. du Caire 67 (M6-langes Maspero II) 1934, 152 ff. (5) P. Berol. inv. 16705 (fifth-sixth century,Hermupolis; 77 entries from &yXLpohAovhrough rvaTrp-aac;full alphabetization),ed. A. Henrichs and W. Muller, Collectanea Papyrologica I (cf. above, n.ii)27 ff. The last is the earliest extant copy of a fully alphabetized lexicon of anytype.


15/28

324 Timothy Renner

Tt3qi MI!'I.*>xOc44

4~i

"IT.. H ~ J I " " " r -a k r i l o ) ? ? , O ?c?-

1 4 A J a rA ~ a s

T -C a?OA, ,o-IN.,W,

FIG. 2. P. Mich.inv. 545Ia

The four lemmata in question in the papyrus presumably survived,whether in editions of Apollonius or in intermediary sources, longenough to be incorporated into either the original Hesychius or theexpanded Hesychius of the tenth century. They had however probablydisappeared from the edition that was the ancestor of the Codex Cois-linianus. Such were the effects on Apollonius of the many centuries of


16/28

ThreeNew Homericaon Papyrus 325recopying,excerpting,and abridgmentto which his compilationwassubjected."5

The formatof the text is of the type most commonlyfound in papy-rus lexica.Lemmataandglossesare writtentogether n a singlecolumn,with the second and any furtherlines of an entry distinguishedfromthe firstline by eisthesis f one letter-width.A spaceof one letter-widthalso separatesglosses from lemmata.leThere are no paragraphi,punc-tuationmarks, accents,or other lectionalsigns. Spellingerrorsfall intotwo generalcategories.The first includesminorvariationsof the typesfound in many papyri (fr. I, 11.8 and 9; fr. 2, col. II, 11.2, 8, 9, 13, andprobablyI5). The secondencompassesomissionsof syllableswhich arepotentiallyserious for the understandingof the text (fr. I, 1. 7; fr. 2,col. II, 1. I2). The errorin fr. 2, 1.8 has been correctedsupralinearlyby the writer, althoughno cancellationof letters has been made. Allothererrors have been left uncorrected.

FRAGMENTI

.YpoAv[/tl?7-stroAvyvOw (gl. I)uwovroAv/ovAog7roAtrrpo7roSm [oA (gl. 2)Aa-rpEr7Wov?vVta5 voCav7roAvKaMprov u[E7a (gl. 3)roAAov Ka,ovy[yov)?a troAvv Kaua-rov trapE{}XyovT[aIo [19 7ovr]pAvEpyov

].[I. 7r: Foot of right upright 7. KacLov pap. 8. yova'ra pap.Ii. ]. [: Horizontal romtop or middle of letter15For discussion of the interrelationships of the various lexicographical com-pilations during this period, see Henrichs, ZPE 7 (1971) 97 ff. (especially I2 ff).On the date of Hesychius see the edition of K. Latte, Hesychii Lexicon I (Copen-

hagen 1953) vii ff.16On the format of ancient lexica on papyrus, see Naoumides (above, n.12),184 ff.


17/28

326 TimothyRennerFRAGMENT

COLUMN I COLUMN II]a rop.[ a (gl. 4)] 7WoOv 1871{}roppEvCO[aL]E7ropVpEL, VopVPLf f (gl. 5)S.............apaaa o v5 KGLKGAXGLVELVEy[E

[T]cLTo 7apaXUU.LY7ro]p&8ALs apa-qva (gl. 6)aA[o ]9 rov 7rpOAaME0atl[

[7rop]vpeos OavaroF.[ iLpe (gl. 7)[Aas] KGL(a)>Ovs[7rop]E{s}) ESwKE (gl. 8)[7rop]T-aKLOaXwr[L (gl.9)15 [2-3]p.gv Saiuat . [[2-3].a[7rop]ovpE E&ctraTev[E (gl. Io)[pcpt]pva[.].. .. ]vw KGTGr[ (gl. I1)20 [2-31.W..... ] avqrpyvv[ (gl. 12)

Col. II I. irop.[:or 7ropT[ 2. About I cm to left of beginning ofline is a small mark like a dash 9. O-qAqtap. 12. KaeOv pap.15. .[: section of low horizontal, perhaps from finial 16. ].: anupright i9. Or ]Etv

Fr. I 1-2 (gl. I). iroAvyvi]/wov A. Henrichs. The word is un-attested as an explanation of 7roAVr/qLrs;ccording to LSJ it occursonly in Plato and Dio Cassius. Also possible is 7roAvbpdi8]1wv. Thelemma may have had the variant form iroAv//rrs, as in Ap. Soph.133.10 rroAvWjrrj"roAf3PovAos,roaA&PovAAEvaaOaLvvdL.vos and inHsch. rr2889 7roAv~w4r'"7roA'foPovAos,oAv`pwv (7roAv/xqra ccurs inOpp. H. 5-5). Suda rr 1987, Phot. s.v., Ba. 345.29, and [Schol. A (inter-lin. gloss, Dindorf) ad 0 355] have 7rroAv'/-rts"7ZroAv'flovAoS.f. Eust.1462.19 (ad y 163).3-5 (gl. 2). So [D. ad a], with the addition of vvverv, i rroMACivE/MrELpovpdrwOv.Similar is [Hsch. rr2934] 6 E7VOWlAA TPErrLEvo,71


18/28

Three New Homerica on Papyrus 327-prwv -ri7v eavro~Stvotav i~5' 'va (o3K ie' 'va Heinsius) KaLpdv.Cf.Eust. 1381.35 (ad a 163). The entry is lacking in Ap. Soph.6-io (gl. 3). The word 7roAv'Epyoss not attested as an explanationelsewhere. Otherwise, Ap. Soph. 133.11 is nearly identical with theentry of the papyrusbut specificallymentionsthe Homericuse of theadjective as an epithet of iron: rdv cra'8pov ETTOETEAK~, r3V tLE7& oAAoOKcq.GTOVEYEV7LEVOV I 7-VvroAvV K"qlaTOV rapexova.Cf.Hsch.7r2870LET TroAAOViaCLXTOvYEyEvr,7OV,IqroAvV Ka~ciTaovt1LvLrapXoV7(7rape'Xwvcodd.).A",ye6'ardvad6lqpov;D. adZ 48] Lpe&roAAMoa(tXTovyevdOlEVOS 7 0 7rOAVKa LaTOVvr7raPEXodLEVOS.therdiscussions,oftenwithsimilarthemes, appear n Eust. 623.62 (ad Z 48), Schol. bT ad Z 48,[Schol. BEHPQ ad 8 718], Schol. Q ad 6 324, [Schol. PV ad 0 io].Ba. 345.26, [Suda r 1986], [Phot. s.v.] give simply 7roAvKcLarov.

Ka()ov: For comparable examples of haplography with omissionof vowel and following consonant see Mayser/Schmoll 1."1.219.y[E]yov(o)ra (y[e]yovara pap.): For a written in place of o see May-ser/Schmoll 1.12.70 f; Gignac I 287 f.7rape{.}Xo.r[a: The spelling may be due to confusion with the im-perfect. The examples of ELwritten instead of E that are given byMayser/Schmoll 1.12.41 f and Gignac I 256 f mostly fall into specialcategorieswhich do not applyhere.Fr. 2, Col. II 1-2 (gl. 4). The lemma for this entry must havebeen 7rdpLsor 7roprLs. The lexicographical sources regard these twowordsfor "calf" (as well as ?rdprac,on which see gl. 9 below),as inter-changeable and undifferentiated n meaning; cf. Ap. Soph. 133.29,Schol. AbT ad P 4, Eust. 534.28 (ad E 162), 0o91.2 (ad P 4), 1625.42(ad t 222), "Zonaras" 1566 s.v. rdp-rtS,EM 684.1, Tz. ad Lyc. 184,D. ad P 4. The etymological theory connecting 7rdp p//7ro'pr7rdpraewith 7ropev'Ea0aand other words from the same root occurs in the

"Zonaras,"EM, and D. passagesas well as in Orion 125.20and Eust.534.28. The last most closely parallelsthe wording of the papyrus:77SE 7r0pTW, a E5SV AAOLS7rOKOPLUtaLKOJS9rpora?, &r oZiv - LaAjvT7)V aptL T7opEvEcTaOLSvva/_L~v17v. ypa4E7Iat SE 7TOTE Kat Xwpls9T7ropL~.Assuming that the writer indented here by mistake,the entry mayhave begun with line i of the papyrus.In that case, Henrichssuggestsrestoring rro'pt[ESgvw't flOdES&]d7rTo) 7('7{L)}rropErdO[atcf. Schol. QVad K 410, Schol. T ad E 162). But if the entry began with an earlierline, rrop.[in line I would be part of the discussion or explanation.Finally, we may comparethe (probablyabridged) entry of Ap. Soph.133.29: 7r' t9S7'v &qa'laAwVKaAVoEV.AEyE?La-7v KaL rofdp-aKa7a &Z


19/28

328 TimothyRenner8' &p' abr lftatv'&i~Sread wA 7er) eEpLrOpKLo Kr rp" [P 4]. Yov7-a86 KGl%7roPLES.

,q8,q{f}: he extra t is frequent during this period. See Mayser/SchmollI.12.Io6 if, especiallyio8; GignacI 185;W. Cr6nert,MemoriaGraecaHerculanensisLeipzig 1903, repr. Hildesheim 1963) 45 f. Foranotherpossible exampleof the same tendency cf. line 9 below.3-6 (gl. 5). Ap. Soph. 133.32 has 7ropOvp7- ropObvpln7rat,rapdaacrcrra"W(S 8' OTE opopvp 7eAayoS" 3 6].Ier-v oiV 7e7roL7ijqq' Aefjog,r popp'pwv. Abbreviated but similar are Hsch. I 3081irop/o'pe"1apaTTerat. bpovrLTE.EavISELt;7T 3086 ropovpeL"LEAaV?EL, ap&TTEL,ropoUvpL'EL;. ad E 16 7rop0Vpr?7L"eAavEL, rapuaact; Ba. 347.6 andPhot. s.v. rropodpevTapaaaeraL. Cf. D. ad 8 427 (= with minor varia-tions Schol. BEPQV ad loc.) 7rodpvp" v flaOeL , 8wavolas StEvoE4roKG KLVEETOGl ETapdaUUETO,WaEp UV1palVEL T7WVS&7WV 7 Kf&O6dvsKLVOIVLEVachEAvEaOL;Schol. A. ad 16.But for parallels o KaAXaIVEtn the papyruswe must turn to Hsch.K 550 KGAXaVELTapda~7E,7ropdvpEL....

K fvOovTapdaaaerat;Eust.964.48 (ad S 16) 7ropcvpevSwE7'TLEAaWVEaOGL,JEVKa KVJpalrOp9vpEOV,KGlItOpovpEos Ocvaeros.&n84E TOp!p0tELVKal KaAXaLVELVEETyraL,a&MaXdoOablVETra.More interesting is "Zonaras" 1171 s.v. KaXaUVELV[Soph. Ant. zo20]' GTr&d0oso-LEpLVaV,KGl(AXt 7 7'ropOV'pa,80EVKaropfv'pELawv TLEpLvayVnearly identical is Orion 79.I). Henrichs pointsout the striking resemblancebetween "Zonaras" (on which lexicon,not compiled by Zonaras, see K. Alpers, RE X A [1972] 732 ff), whouses the Homericgloss to illustrate he metaphoricalusageof KaAXalVELVin Sophocles, and the papyrus,which cites the Sophocleanword as aparallel o Homericusage. Ap. Soph. does in fact quote passagesfromSophocles on occasion (91.34, 148.zo). See also glosses 7 and io below.7-10 (gl. 6). The fullest lexicographicaldiscussionof the etymology

of 7r'dps(xA s Ap. Soph. 133-34 rdp80sa"S oroaro0o/ A~yovros 7roTElEv ; 70TO) TOLXElov, r7dpaAWLS1 LEyas vs" [L 457], ro0 8E &&700 a, "rapaAv oqL;vOErLVXwv" [r' 17]. 24!ywov(fr. io9 Neitzel)SE@O-8&a90pELlP 7rdp8&ALVEVyap i-T&paEVLKdV,7rxp8&AtV E 7TI)VGI)AEUXV,(XLTOv/PVE ro 700 7rpoaAAEirOa,V 8E a7ro To 7rapa&AAXMa. Pj7Tov 8S7 KOLVWSlVETlral eAyV, "t Oowvo op8aAlwV(TE)AVKWV'l a 7rAOVtaL[N 103], WE dEl0r 0 (C'0V 8t' 70To 0, 8E E 7rS 80op. &08 - a. Ascorrectedsupralinearly,he gloss of our papyrusneatlysummarizes he

17 The spelling roop?%pELswas favored by Zenodotus, accordingto Schol.

A'mTIIad H 16. But its appearance in our papyrus and in many of the lexico-graphical parallels is adequately accounted for by itacism and assimilation tothe indicative.


20/28

ThreeNew Homerica n Papyrus 329commentsattributedby Ap. Soph. specifically o Apion.The latter,theauthorof a Homericlexicon and probablya pupil of Apollonius'father(rather han of Apolloniushimself,as aprobablyuntrustworthyancienttradition proclaims [cf. RE II I, 135 f]), was interested in fancifuletymologies.Hsch. rr3009 has a discussionof 7rdp'aALnd 7rcdpALsthat is identicalin content with Apion's as given in Ap. Soph. and thepapyrus, except that the two infinitives are misspelled as vrpoeA&eTaL(cf. pap.) and rrapaAoeat.The same explanation,havingthe infinitivesspelled correctly,turns up in Eust. 1092.39 (ad P 20).18 For similardiscussions based on etymology see "Zonaras" 1562, Et. Gen. (AB)s.v. 7rdpaALS,. ad P 20, Eust. 1251.51 (ad 0 573), 789-39 (ad K 29),922.53 (ad N 103), Ep. Hornm.P 173.28, Et. Gud. 452-49-56, EM652.27, Phot. s.v. VrdpSaAw,chol. T"1ad r 17, Schol. AT"'ad K 29.Some of these try to explain the nouns by sexual differentiationandothersby distinguishingthe living animalfrom its skin (Ep. Hornm.O356.11 does some of each); many incorporate,or are based on, in-correctspellingsof the infinitivesgiven by Apion. Cf. also Schol. Alntad N Io3, Schol. AImbTad P 20, Schol. AImbTT" d 0 573.OrlA(ELG)0?rA7rLap.): Perhaps the exemplar had ,7rA-7aand thewriter of our papyrus omitted the final syllable (which would be inkeepingwith his errors n line 12 below and in fr. I, line 7) but addedan extrat to the rqas in line 2 above. For rqwrittenin placeof ELbeforevowels see Mayser/Schmoll I.12.58 ff; Gignac I 240 ff. A spellingOnrAnan the exemplar would seem less probable (Mayser/SchmollI.12.1o3). On the otherhand, perhapsthe falseformOrAtqwas intendedby the writer(deviationsof this type involving adjectives n -vs, -ELa,-v are discussedin Mayser1.2.52).

11-12 (gl. 7). So Ap. Soph. 134-7,Hsch. 7r3084 (with the additionofKilcapax&SXrs).M 684.10o, chol. AbT ad E 83, Schol. bT ad S 16,Eust. 524.42 (adE 83), 964-43(adS 16)all eitheruse /~Aasto describea 7rop vpEOsOdvwrosr connect the phrasewith a verb related to theadjective(cf. "Zonaras" 1568 s.v. 7ropovpEov).See also the referencescited for glosses 5 and Io.

KaL(


21/28

330 Timothy Rennero w7rapaX "rdp' v ' eE ipatra Avypd" [Z 168]. Cf. also theentry in manuscript V of EM that is reported in Gaisford's app. crit. on684.18.

[7rop]e{s}: Mayser/Schmoll I.I2.182 f lists a number of examples ofconfusion between final -s and final -v. Many are simple mistakes orarise from nearby case endings; but Gignac I 131 f suggests a phono-logical explanation of some of them.1914-16 (gl. 9). Probably the entry should be restored [7rdop]TCKL'/LocXoWL.Tr[opTr /6 op]p-qv, ,tpaAwv(read &84laLs) ~ [04/Ahe]a. Theerror in case ending is probably due to the -v ending the precedingword; cf. Mayser/Schmoll I.12.182 f. The resulting count of five lettersin the lacuna to the right in line 14 is a bit high but possible, sincecalculation in other lines shows that that edge of the column may nothave been even.The differentiation of r7dp-raend 8t4/LaA according to sex is ascribedto Aristophanes of Byzantium (Codex Athous ap. E. Miller, Milangesde litterature grecque [Paris 1868] 430 = H. Erbse, Lexica GraecaMinora [Hildesheim 1965] 276). Ap. Soph. lacks an entry on rTidpeas a lemma, although 133.29 (quoted above in connection with gloss 4)does discuss the word along with its synonyms radprTSand raodpr.Hsch.7T 3071 has rPpTaKL'LdGX. -atdaE; cf. 3073opa~ "rT pp v oSg.TwS 8&aLaLv,,AAOLEoyvdY, of 84

% ov. On rodpP'raee also Schol.AbT ad P 4, Eust. 534.28 (ad E 162), 1091.2 (ad P 4), 1625.42 (ad t 222),Hsch. 7r 3068 and 3070, D. ad P 4.17-18 (gl. io). The entry is lacking in Ap. Soph., although 133.32concerns the lemma 7ropbl'p (above, gl. 5). There are no parallels tothe use of a-,aELVto explain 7roptp?',Ev. However, the remainder ofthe gloss in the papyrus can be supplemented by comparison with D.ad 0P551 ~dp

uvpeS. KeaT&dOog E)LEPlVPaKati

EVOE-TO.For 1rTdp4vpE

explained by a form of LEptpuVVee also Hsch. r 3082, Schol. Aint ad 0P551, Schol. T ad 8 572 (cf. "Zonaras" 1171 s.V. KahXAL'VELV, 566 s.v.Vropfvpa, 1570 S.V. lropf ,pw, Schol. bT ad E 16, Et. Gen. [AB] s.v.wop v'p , EM 684.1o, Schol. Ap. Rhod. 1.461, 2.548). The AEELr'O/pcL7PLKCLattK UTO'XEOVof Codex Urbinas 157 and Codex Selestadi-ensis 107 give the following explanations: Cod. Selest. 126v 7TodpbvpE-eveOvLpe7ro;Cod. Urb. 210v rrop4vpEEv[sic]" EOvLEEiro[sic] (from notessupplied by A. Henrichs; see his references in ZPE 7 [1971] O10, n.I3,and I 18 to these two largely unedited manuscripts). Further discussions

19 On the other hand, wropEather than wopev may have been in the writer'sexemplar. For mistakes involving the addition of final -3 to a vowel see Mayser/Schmoll I.12.183; Gignac I 125 f.


22/28

Three New Homerica on Papyrus 331of 7rr'dpvpe re given in Schol. BEPQV ad 8 427, Eust. 1251.3 (ad 0551), D. ad 8 572 = ad K 309, Ep. Hornm.P 28.34; and see glosses 5and 7 above.

19-20 (gl. II). [ropual]vELv"Ka-a[aUKEv]ELtVJ. S. Rusten, comparingHsch. 7r 3056. Ap. Soph. has no such entry.21 (gl. II). The lemma was probably w?rorLs.he simple roag"3&v(4pappears in Ba. 347.8 and Phot. s.v. Ap. Soph. 134.8 has 7rado' "rroTEFE r7mSTOUEWS,roT- TT70O(XT VodLov v8pds;cf. Apion 99.9 iroa- TB7rtdVtLEVov.K d &v'4p.The gloss in the papyrus must have included anexplanation of rroado ("husband") based on a supposed connectionwith ~aITod'"drink") and wTort~ELV"give to drink," "irrigate"). Suchan explanation forms part of a rather lengthy etymological discussionin EM 149.41 p0lV" trapa& O posw, 7rorlSw . . d y&p vi7pP yvva leTCapsEVEL. P8W,PuoW, cPPo)V, XPSWV -iV 01AIELUV. Kal yd(p &(X TOVTOTo'drLAyETa &vo p Tcapcp woITT7~o yot Vr.... Closely similarpassages are to be found in EM 149-50, 684.19, Et. Gud. 207.1, 207.6,Eust. I384.59 (ad a I5). Cf. also Hsch. r 3104-3107. The entry in ourpapyrus must have continued over into at least the first line of thefollowing column and must have read approximately as follows:[1TdoS]'

&v7p"vv[catKtCy&p TT071rELor pSEL)], perhaps followed by

further comments.

3. ANTHOLOGYRSUMMARY-WITH-QUOTATIONSFIliad E--TP. Mich. inv. 4832c; 11.5 x 8.6 cm; late second orfirst centuryB.c.;KaranisPeppered with worm-holes, this fragment was obtained during the

1926 season at Karanis. The hand establishes it as one of the earliesttexts from that site.s2 Two columns of writing parallel to the fibers20 The excavation number is 26/BI3F/A. The probable date of the papyrusmakes it an unusual one for Karanis, where the great bulk of the documentsdate from the Roman period and most of these are later than the first centuryA.D. The ostraca and documents occasionally bear late Ptolemaic dates, andseveral have been assigned to as far back as the second century B.c. (see O. Mich.).For the probable foundation of the town under the early Ptolemies as well astwo inscriptions from the second and first centuries B.C., see E. Bernand,Recueil des inscriptionsgrecquesdu Fayoum I (Leiden 1975) 164 ff.I have not made an exhaustive survey, but probably our papyrus no. 3 is theearliest literary piece to come from Karanis. Together with no. 2, the Homericlexicon, it adds to our information about the modest degree of interest inliterature there. Besides more than zo texts of Homer, papyri of Hesiod's Th.


23/28

332 Timothy Rennersurvive; but both are incomplete at the top and the first has lost fromone half to three quarters of its width on the left. The writer was hardlya raw beginner. However, his shaky rendering of a characteristic late-Ptolemaic book hand and his lack of success in producing uniform linesand column edges have resulted in a text that is not particularly attrac-tive visually.21 The back of the papyrus is blank.The extant part of column I begins with five verses quoted from thecatalogue of Nereids who accompany Thetis to visit Achilles; presum-ably the papyrus originally contained the entire catalogue (E239-49)from the beginning. The remainder of the column, in prose, apparentlysummarizes the following events through Thetis' visit and her departurefor Olympus to procure arms for her son. The upper part of column IImust have dealt with the making of the arms (concluding 2) and Thetis'return with them to the camp of the Achaeans (T i if). The first pre-served lines of column II show that the compiler of the text has onceagain turned to quotation, this time of two verses (T 38 f) describingand Op. (cf. BASP 3 [1966] 65 ff), Demosthenes, Isocrates, possibly Xenophon,and Chariton are listed in Pack2 as coming from the site (see nos. 243, 316, 1246,1256, 1552). Among as yet unpublished papyri from Karanis in the Michigancollection are fragments of Plato's Phaedo and a grammatical treatise, datableto the first or second century A.D. A scribe from Karanis inserted a reference toCallimachus in a tax roll which he was copying (see H. C. Youtie, Proc. TwelfthIntern. Congr. of Papyrology [Toronto 1970] 549 ff = Scriptiunculae II [Amster-dam 1973] Io39 ff), but of course this is an eccentricity. Finally, of specialinterest in connection with our Homeric lexicon and anthology of the Iliad isyet another literary papyrus from Karanis, the Life of Homer by Alcidamas(Pack2 76; second-third century A.D.). Taken together with the actual texts ofthe Iliad and Odyssey, these three Homerica demonstrate that rather extensivestudy of the poems was pursued there, although much of this activity must havebeen associated with the schoolroom.

21 Similar but in some respects cruder is P. Ryl. I 30. Similar but more orna-mented and regular are PSI IX o1092 = Norsa, La scrittura letteraria grecaTav. 8; Turner, Greek Manuscripts Pap. 55. All three have been assigned to thefirst century B.C. Also comparable on many points are Roberts, Greek LiteraryHands Pap. 8a (99 B.C.); 8b (30-29 B.c.); 9c (late first century B.C.); Seider II,Pap. io (second century B.C.); Pap. 15 (first century B.C.); Turner, Pap. 12(second century B.C.). An oddity in the Michigan papyrus is the second w inII.xo, which is shaped like (). Perhaps this is due simply to the whim of thewriter on this occasion. I do not know of a parallel to it in this type of hand,but it could be an adaptation of one of the earlier forms of the letter which are inuse in the third and second centuries, especially in documentary hands. As amark of a certain pretentiousness in the writer, despite the clumsiness in writingand format, we may note the "filler " mark at the end of I.14. The intercolumnarspace fluctuates between o.8 and 1.5 cm. The lower margin of up to 2.2 cmindicates a fairly generous use of space.


24/28

ThreeNew Homerica n Papyrus 333Thetis' preservation of Patroclus' body with nectar and ambrosia. Theprose summary then picks up the narrative of the subsequent assemblyof the Achaeans, including Agamemnon's bestowal of the promisedgifts on Achilles and his oath not to have slept with Briseis. For theoath, the compiler incorporates a single Homeric verse (on which seethe commentary). The column concludes in prose with the end of theassembly, Briseis' lament over Patroclus' body, and subsequent eventswhich are perhaps those of T 303 ff but which cannot be identified withcertainty because of the fragmentary nature of line 17. Presumably thenarrative continued into a subsequent column.This type of composition, having noteworthy or memorable Homericpassages or verses spliced together with rather simple prose thatrapidly summarizes the intervening narrative, appears also in threeother papyrus fragments which are the object of a recent study byG. Nachtergael, "Fragments d' anthologies homeriques," in Chroniqued'tgypte 46 (1971) 344 ff. These papyri are P. Strasb. inv. 2374 =Pack2 1185, ed. N. Lewis, Atudes de Papyrologie 3 (1936) 46 ff (thirdcentury B.C.,provenance unknown; on Z); P. Vindob. gr. inv. 26740 =Pack2 791, ed. H. Oellacher, ibid. 4 (1938) 133 ff (second century A.D.,Soknopaiou Nesos; on Z); P. Hamb. II 136 = Pack2 633 (early thirdcentury B.C.,provenance unknown; on B). Together with the Michigantext, these fragmentary anthologies bear abundant witness to the con-tinued popularity through five centuries in Greco-Roman Egypt of atype of study aid which had probably been in use in the schoolroomsince at least Plato's day.22In the papyrus, blank spaces are employed in 1.6 to separate twoquoted verses, in 11.5 to indicate a break within the prose narrative,and in II.II before the sentence which leads into the verse describingthe oath. Probably this device was used more extensively in the longquotation in column I. There are no marks indicating punctuation andno accents or other lectional signs. Iota-adscript is written. The onlydiscernible spelling errors are minor: one in 11.17 left uncorrected andanother in 11.8 which the writer seems to have caught.

COLUMN[ ,WpLS,KL Ha17vo].).KacL

(2 45)[ayaKAEL-rT 1ACIrELCNyLE]pTrI?r]E (46)[Kat AObEv877S a~L aA]ALavcnaaa

22 Cf. Nachtergael,347.


25/28

334 TimothyRenner[Ev0a SE qv KAvELEV7Ia]yELpa rE (47)5 [Kat Iavaota Matpa] Kta QpELOvL (48)[a EV7rhoKaoLS 7rAa0]va x aAAa (49)[0 at Ka-a /3EVOOS aAo]. NTqp)LSE5[rqav ca.12 ].oqt4tVT-KE[ Ca. Is ].EL TWLAXLAI0 [AEL ca. 12 ]p. (EavoveLE[ ca. 13 ]..[.].Ov ora E[ ca. i3 ]... ETaL, ' E Tao[ ca. 12 ]wEWibaua 7rpos- Tov[ ca. '4 ]Vq8cs-Erov>

9. ] : Perhaps A io. Before0, either o or E ~Ea: After E, perhaps anarrow letter such as L lost in gap i i. ].. [: foot of diagonal, thenupright and low ink Iz. ]...: in middle of area, parts of two uprightsCOLUMNI

Ha][rpoKAWL av-ra]lPf poL1r7v Ka~ VEK] (T 38)[rap] .[pvOpov n]ra.[e]. K4[a-ra] (39)[pt]y[wY wv]a Ot XpWS EIj[wE][80o].4[ Ki]c.]rov7OTOL~rl[a]5 [Caa]Xowptra AXLAAEvS 8E[a]uyayt. ravras

rovs AXat[o]vS9f9 EKKAqrltCavKaLVTOL

EqfTat27r jqtrl8soS a7roppqavVAyarLEpwLVOvSe av-rwoIO arTO&tWootLrarv-ra ra 8wpacaa7l7 Bpirqt[s]a OtLVVELEFLI7TroTET77S EVV7s EIrTflp (T 176)WLEavL1E

7wy.VtKaLOVTws9AveraLr EKKA[q]OULa

I5 Bptopt -SOeraa u 7"0rorvo Ha7TpKAOvUWata OpVL Troya. rqeovoa.. . a.o8. e is written in margin, apparently to correct {,} (of which the rightupright survives) I7. 8.: perhaps either SEor 8,7 ...: an uprightfollowed by a gap; then top of rounded (?) letter and another uprighta: or S; next letter has part of an upright .a: bits from base of anupright (prob. t), then top and base of a with perhaps ink from left side


26/28

Three New Homerica on Papyrus 335

? 0 0

, K -4 01-it*- T77?:: A

: . W T

4':iAWNxtNOY S

en00C

,-En6cdi-.


27/28

336 Timothy RennerCol. I 1-8. The catalogue of Nereids (E 39-49) was athetizedby Aristarchus; the Argolic edition of the Iliad did not contain it (Schol.A ad E 39; cf. T ad E 38).6. ApaO]vLa:The spelling in -vta rather than -ELa could be due toerror on the part of the writer by analogy with the preceding OpELOvLa.However, some manuscripts do contain this variant, and it may havebeen present in the text of Homer that the writer had in front of him.T. W. Allen's edition of the Iliad (Oxford 1931) gives the followingdata (ad 2 48): -Eta Bm2Bm5L6L7L15L17L18L19L20Mo105P2p3p4p6p15p21

PaPeU9U13V1V4V5V6V15V16V20V23V26Vi2Vi5Eu;VELaP12V2; -vwa cet.Eust. 130.56, discussing the etymologies of ApdLOEand 'QpdOEvLa,disapproves of 2ApdOva .. .'v o03 KahAW WVES..'dh.r'AUav ypdaLVrpo 4olzo7o-ryra0ro QpEclOav.8-io. E.g., (sc. 9E'r1) 87%-rvvhv 6r]oLTUzt/Uzv'KE[raal aVEAOOoV'a(sc. EK 7T- OaAo'ouu)dcpt]AE-i- AXLA[AEFalthough a bit long for thelacuna in 9). Cf. E 51 ff.io. Perhaps -r]pobE[Z]a (see app. crit.) or a word from the same root.Against this would be the fact that in the standard text of the IliadThetis touches upon the birth and upbringing of Achilles mainly in, 51 if, before she comes from the sea to comfort him. But this causesa problem only if the gist of my suggested supplement to lines 8-io iscorrect (and E 95 f might also have suggested the theme to the writerof our text).

1i. E.g., rap'C'Hb]a[a]irov &TrAa.Cf. E 136 if. (&ra might also referto the armor which Hector stripped from Patroclus' body, sinceAchilles mentions this in verses 82 ff.)12-14. E.g., (sc. O1ELt) SE -ral [&8EAea`3..] rr7Tzaao'a rp%3rov

[rra'rpa ... (cf. E7140 ff). But this does not seem to go well with thefollowing]vly8ELorov,

a sequence in which one would most naturallysee " (sc. 9E9'LS)8' ESr7v ["OAvirTovor E'l 7-r [70o3' "Hacrov oKOV)rropEerVEa or the like (cf. 2 146 if).

Col. II 4-5. K]aL0-rovro r7Torl[aua]XOpt57raL: This sentencemay have been purely the invention of the author of the anthology,without any particular basis in the text of the poem which he had infront of him. However, the detail was more likely present in his text,since several medieval manuscripts have an "extra" verse T 39a, onwhich Allen's apparatus reports as follows: 39a F& v8P' c~p'vopea'&drrE7Efl0/3~r apyvpdrroEaort. yp. T, vide Schol.; habent Bm8GeL2L12Le'M4P7P21Q4U3U13V20V25.


28/28

ThreeNew Homericaon Papyrus 3375-10. Compare the corresponding summary (hypothesis) which pre-cedes the scholia on T in D. and Eust.: avvayayWv El, EKKA-o'aOV70ro&EAAnvas0' ?pws* &7rp'pp'qcnv 7TOLELIatL -4 77'VL8OS.KAtAcpflc~Va- -a -paaI p alT vyapcelEv 7rrTEUXEro.I1-13. In the text of the poem the verse T 176, embodying thesubstance of the oath in a single line, is actually spoken by Odysseus.This occurs well before the actual swearing of the oath by Agamemnon(which is done in verses 258 ff with elaboration and without repeatingverse 176) and the breaking up of the assembly and lament of Briseiswhich follow closely on the swearing. But it would be rash to assumethat the text of the Iliad before the writer of our anthology necessarilyrepeated 176 as part of the actual swearing or through some otherdeviation caused him to put this verse in Agamemnon's mouth. WhilePtolemaic papyri do show some eccentricities and "plus-verses" in thetext of Homer, verse T 176 (= 1133) is a concise and easily rememberedstatement of the oath which on account of these very qualities might wellrecommend itself for inclusion in an anthology.16-17. Probably 7-v (sc. AXLAAEM)U ?glovuav (read jlweuav)...,the verses referred to being T 303 if, where the y'povmrS gXOauv begAchilles to eat before going to battle (cf. Odysseus' remarks in T 155 if).Assuming a letter lost after a, we might restore 7-v / 8U '7~eJwav o&[p]LUErs[7)pvC-c-redEw.. On the spelling error in the verb cf. Mayser/Schmoll 1.12.76 f; Gignac I 210 f.

MONTCLAIR STATE COLLEGE

Three New Homerica on Papyrus - Renner

Documents

Transcript of Three New Homerica on Papyrus - Renner