RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 Some Passages In Plato R. Renehan I Meno 98A rovro ~'eariv, OJ Mtvwv haipe, dwiflV'1aze;, we; ev roie; 7Cpoa­ (Jev ~fliv Wfl0).,0Y'1ral. w F: om. BTW "OJ Mivwv eraipe. "'Q is not normally omitted (as it is here by BTW) unless the tone is somewhat peremptory (Kuhner-Gerth, I, 48) .... In our passage Thompson reads OJ but follows Naber in excising Mtvwv as 'a gloss upon eraipe which has extruded OJ'. Possibly the ommission of OJ can also make for impressiveness: d. Crito 52a, where B omits it: if so, this might justify its omission here. For other possible instances of its omission, d. Hipp. Ma. 289b, Soph. 220d, Lach. 198a." R. S. Bluck ad loco Bluck (as also Burnet and the Bude editors) prints OJ Mtvwv haipe on the au­ thority of F,l but it is clear from his note that he had doubts. Actually, Bluck's approach to the problem misses the main point, since every parallel cited in his note is an example of w expressed or omitted with an otherwise unmodified proper name. 2 What is wanted is some account of the ways in which Plato employs the vocative eraipe, and that I propose to give. First, by far the com­ monest use is the simple w haipe, with no proper name expressed; this occurs over seventy times in Plato.3 The evidence thus is overwhelming that Plato's normal, and frequent, practice was to write OJ eraipe, not haipe. The next commonest locution is w rpO.,e eraipe, again with no proper name expressed. Brandwood lists nine examples;4 rpi).,e haipe is not found in Plato. In addition, there are three other Platonic occurrences of the word (not counting 1 For F see E. R. Dodds, Plato Gorgias (Oxford 1959) 41-44, and R. S. Bluck, Plato's Meno (Cambridge 1964) 135-40. 2 E.g. Soph. 220n cO 8eairTfre W Stab.: 8ea(rTfre BT. 3 The passages can be found in Leonard Brandwood, A Word Index to Plato (Leeds 1976) S.v. eraipe. 4 Supra n.3. The passages are Euthphr. Sc, Grg. 482A, Hp.Ma. 296A, Ly. 213B, Phd. 91B, Resp. 450n, 459B, 562A, 607E. 371 RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 372 SOME PASSAGES IN PLATO Meno 98A), all of them unique types for Plato: OJ eraipe ;.t vvre (Meno 94E2); OJ rplAe eraipe Kpirwv (Crito 540); OJ eraipe LlJ1,J1,la re Kai KiP"C; (Phaedo 82c). I return now to Meno 98A. Both OJ Mivwv eraipe and Mivwv eraipe are singular expressions;5 that is no argument against them. I have just given several unique forms of vocative address in Plato; here are some others: OJ rclv (Ap. 25c); OJ J1,iAe (Tht. 178E); OJ rplAor"c; (Phdr. 2280); OJ rplAOUJ1,eVe (Symp. 201c); OJ rpiPl(JU (Phdr. 2380); W IIpdnapxe rplAe (Phlb. 530); W 1tai rplAe (Soph. 230c).6 Clearly, the fact that a vocative phrase happens to occur once only in Plato is, in itself, insufficient reason for questioning the phrase. In the present case, the position of eraipe after the proper name (with or without w) is most unusual, a phenomenon which Bluck, to judge from his silence, does not seem to have adequately appreciated. But the position of rplAe in w IIpdnapxe rpl).e and w 1tai rplAe (supra) parallels the position of erafpe here and may be taken as a reasonable guarantee of its soundness. 7 This brings us back to the elusive W. Is it possible to determine with any assurance whe­ ther Plato wrote w in Meno 98A? I think so. Read w Mivwv eraipe. In all of Plato there is no example of eraipe, alone or in combina­ tion, used without W. Mivwv eraipe is a singularity which I am not prepared to admit,S and Philebus 530 w IJpdnapXe rplAe, an exact parallel to w Mivwv erafpe, ought to tip the scale. The vocative wyaBi (sometimes written w ayaBi with scriptio plena in the MSS.) occasionally presents comparable difficulties. Bury prints Symposium 189A as follows: ... Kai rov 'Epvc}J1,axov, 'DyaBi, rpaVal, [.itpl(JrorpaVeC;,] opa rl 1tOleiC; . ... Here is Bury's com- ment ad loc.: "[.itpl(JrorpaVeC;]. I follow Sauppe and Hug in regard­ ing the proper name as a gloss on wyaBi: as a rule, wyaBi stands alone." As in Meno 98A, so here the spectre of proper name as 5 The suggestion that Mtv(JJV be deleted as "a gloss upon b:afpe" is fanciful and should be dismissed. Even were one to consider deletion, the obvious procedure would be to assume that, in OJ Mtv(JJV b:aipe, the vocatives Mtv(JJV and erafpe are variant readings which have been conflated in the text. (Why would erafpe require a 'gloss' here? The passage occurs in the middle of a long conversation between Meno and Socrates.) In that case the choice would be between OJ Mtv(JJV and OJ {raipe; there is no means of deciding. Should any find this solution attractive, let them toss a coin and be done with it. 6 I take these examples from my Studies in Greek Texts (Hypomnemata 43 [1976]) 125 q.v. 7 In Studies (supra n.6) I suggested that this postposition of,f) .. e was perhaps "a manner­ ism of Plato's late style." erafpe in Meno 98A shows that the qualification 'late' should be removed. 8 In rejecting the 'singular' Mtv(JJV b:aipe after my remarks in defence of singular expres­ sions, I am guilty of no inconsistency: both Mtv(JJV erafpe and OJ Mtv(JJV tr:afpe are singular phrases. In such a situation one should choose the expression which agrees more closely with the author's style as otherwise known. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 R.RENEHAN 373 'gloss' hovers over us. And the reason given in justification? "As a rule, wyaBi stands alone." The facts are-my debt to Brandwood's Word Index to Plato will be obvious-that (1) wyaBi (w ayaBi) stands alone about forty times in Plato, and (2) wyaBi (w ayaBi) followed by a proper name occurs a full seven times (not counting this passage).9 So much for the rule. There are no grounds for tampering with 54.purrorpaVer;. Consider next Protagoras 311A: ... Kai eyro einov' "Mr,nw, ayaBe, 8K8la8 iWf..l8V, npcp yap 8a1:lV, aAAa &i5po e~avaa7:(iJf..l8v ei~ !17V aVAr,v ... " f.ainw ayaBe BTW: J1.r,nw wyaBe corr. Coisl.: J1.r,nw ye, w 'yaBe Hirschig: J1.r,nw y', wyaBe Cobet (sed. mox eKelae iWJ1.ev) Both Burnet (whose apparatus criticus I have reproduced) and the Bude editors, Croiset and Bodin, print the reading of BTW as above, {lrrltW, aya8i.10 These editions appeared, respectively, in 1903 and 1955. But already in 1893 J. and A. M. Adam, in their edition of the Protagoras, had printed {lr,nw, wya8i with the re­ mark: "The MSS have {lr,nw ayaBi: probably the archetype had {lr,nwya8i, by a natural mistake. Co bet rejects eKelUe IW{leV, read­ ing f.1r,nw Ye, on the ground that with {l1jnw ye the Greek idiom does not repeat the verb: but there is no proof that the verb could not be expressed with {l1jnw (without ye)." This is basically correct. Nowhere else in Plato does ayaBi occur without w; we should be reluctant to admit the singularity here, where there is such an ob­ vious mechanical cause of corruption. The Adamses are probably not correct in positing an original {lr,nwya8i; the corruption is slightly easier if we assume for the archetype {l1'/nwwaya8e with scriptio plena (compare above). Read then f.111nw, (w) ayaBi (rec­ ognizing, of course, that Plato may have pronounced this wyaOi). II Phaedo 89D if re yap {lluavBpwnfa Bw5veraz BK rov urpobpa !lvi mu­ revual liVeV rixvl1C;, Kai r,yr,uauBal navuinaui Ye aAl1Brj 9 era. 401s, 4280, 436c; Grg. 511s; Leg. 81lc; Phdr. 243c; Resp. 4230. 10 The apparatus criticus of the Bude edition differs slightly from that of the Oxford text: "araOi BTW: r' m 'raOi Hermann wraOi Vatic. 1029 (forsan recte)." The main point, that the principal MSS. have draOI!, and not wyaOI!, is not in dispute. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 374 SOME PASSAGES IN PLATO eival Kai 0"1'1] Kai 1tlurov roy av{}pwnov, enelra oAiyov vurepov eopeiv rovrov novrJpov re Kai a1tlurov, Kai aV{}l(; lrepov' Kai ora v rovro nOAAUKl~ nu{}y n~ KrA.. Read possibly ... Kai av{},~ (lrepov Kai) lrepov' Kai ... ? The same idiom, in a comparable context, occurs on the next page, 90B: bteu5av rl~ 1tluuvuy AOycp nvi dArJ{}ei eival avev r1]~ nepi rov~ AOyOV~ rexv,,~, Kanelra OAlYOV vurepov avrQ) bO(,1J 'l'evb~~ eival ... Kai aV{}l~ lrepo~ Kai lrepoe;. Note the verbal echoes. The same usage occurs in Menander fro 656.8 Koerte, erepav neplf.1£ival xdrepav rpIKVjliav. Compare the similar idiom at Apology 27B, ~ aAAa Kai aAAa {}opvpeirw (where see Burnet for further ex­ amples). With lrepov Kai lrepov, followed by nOAAUKle; in the next clause, compare Xenophon Anab. 1.5 .12 aAAo~ be ... Kai aA­ AO~, ei'l'a noAAof. III Phaedo 99E-IOOA fuwe; jlev ovv cP eiKU(W rponov rlva OVK eOlKev' 013 yap nuvv uvyxwpro roy ev [ro~] AOyOl~ UKOnOVjlevov ra ovra BV eiKoUljlaAAOV uKoneiv ;; roy BV [roi~] lpyOl~. rOle; AOYOle; BW: rOle; om. T Stob. rOle; epYOle; B Stob.: rOle; om. TW So prints Burnet; Robin (who wrongly reports Stobaeus as omit­ ting the article roi~ in both places) agrees, giving in his text BV AOYOl~ and BV lpyOl~ respectively. The oldest testimony for this passage, one which goes back practically to the Academy of Plato himself, seems to have been overlooked in constituting the text, Aristotle, Metaphysics 987b31-32 (discussing Plato): ... Kai 11 rrov eibrov eiuaywy~ b,a r~v BV rOI~ AOyOl~ Byevero uKe'l'Zv (of yap nporepol blaAeKrlKq~ 013 jlereixov) . ... Ross ad 1oc. observes " ... the Platonists are called ol BV roi~ AOyOl~ in 8 1050b35. The phrase used here is pretty clearly a reminiscence of Phaedo 100A, where roy BV roi~ AOYOl~ UKOnOlJjleVOV ra ovra, 'one who studies things by the method of definitions', is Socrates' description of his own method." I agree with Ross that the Aristotle passage is a reminiscence of the Phaedo, but even if it is not, it is a cogent parallel for BV rOle; AOyOl~ (as is Metaph. 1050b35). Compare fur­ ther this very passage of the Phaedo, 99E: el~ rov~ AOyov~ Kararpv- RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 R. RENEHAN 375 yona ev ba;fvou:; aKOnelv. Future editors should print ... ev roie; AOYOle; ... ev roie; [PYOle; here. IV Phaedo 112c orav re ouv vnoxwprW1) ro (J~wp de; rov ronov rov ~" Kdrw KaAOVj.levOv, roie; Kar' eKelVa ra pevj.lara [~la] uic; y#e; dapel re Kai nA'1Poi aura wanep ol enavrAovnee;' orav re au eKel(}ev j.lev o.noA{n1), ~evpo ~e opj.lr,(1), ra ev(}dJt; nA'1Poi au(},e; KrA. bUI ante nk yije; om. Stob., del. Burnet A difficult passage. Burnet renders rOle; Kar' ... dapel "the streams flow into the regions on the further side of the earth," construing ra pevj.lara as subject and taking rOle; Kar' eKelva r#e; y#e; together. Hackforth rightly pronounces this impossible; he ex­ plains the passage thus: "I retain ~ld before r#e; y#e;, and take the literal meaning to be 'it flows through the earth into the beds of the rivers of that region (hemisphere)'. Plato writes rOle; Kar' eKelva ra pevj.lara rather than eKeivOle; rOle; pevj.laaz simply because the pevfJara do not exist until the water fills (or refills) their beds" (Plato's Phaedo 178 n.S). This is nearer the mark, but fails to ex­ plain the dative roie;; as has long been recognized, roie; Kar' eKeiva ... e/apei is very questionable Greek for "flows into the parts about those streams." de; c. acc. would have been expected. Com­ pare immediately above, 112B: orav de; ro br' BKelva r#e; y#e; op­ j.lr,aIJ Kai orav de; ro bri rd~e KrA. Accordingly, Wyttenbach con­ jectured rore for roie; here and Ast, followed by Archer-Hind, deleted the word. Neither proposal convinces. There is a simpler remedy to hand: OLaV re OVV vnoxwptja1) ro iJbwp de; rov ronov rov br, Karw KaAor5J1eVOV, (tv) rofe; Kar' BKelVa ra per5J1ara bla riie; yij~ dapef re Kai nAl1Pof aura KrA. The periphrastic BV rOle; Kar' eKBlva ra pBVf.Lara may be compared to de; ro en' eKeiva and de; ro eni rdbe in 112B. Archer-Hind (ad lac.) has raised a further objection to the trans­ mitted text: "Mr. Cope translates 'it flows through the earth to the neighborhood of those streams and fills them, as it were by a RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 376 SOME PASSAGES IN PLATO pump'. But surely t5,a rr;r.; yr;r.; describes the progress of the water after it has entered the channels: it would be a strange expression to apply to its surging up and down Tartarus." This objection seems quite valid to me; the insertion of (tv) before rofr.; has the further advantage of removing all difficulty on this score .. The corruption is of the easiest sort; uncial f N dropped out after ON. (The omission of t5Ia in Stobaeus may be similarly explained: L1IA fell out after AT A .) v Symposium 176B clKovaavra ovv avrwv e({Jl1 'Epvt;ip,axov rov :4KOVp,eVOV ·H KaAwr.;, ({JavaI, Aeyere. Kai erl ivor.; t5eop,al vp,wv oxovaal nwr.; exel npor.; ro eppwu(}al niVtlV :4 ya(}wv. Ovt5ap,wr.;, ({Javal, ovt5' avror.; eppwp,al. J4 yaOwv: :4 yaOwv (or.;) Vahlen "I still require to hear from one of you how Agathon is disposed with regard to engaging in drink." "I'm not up to it either," said Agathon. Difficulties have been rightly felt about this sequence. What is the point of asking a third party about Agathon's condition rather than Agathon himself, who is present? What is the purpose of the emphatic, and unspecified, ivor.; •.. vp,wv (stronger than, and dis­ tinct from, rlvor.; ... vp,wv)? Vahlen's :4 ya(}wv (or.;) , printed by Bur­ net, provides a smooth thought-sequence at the cost of an intoler­ able word-order. Had the MSS given :4ya(}wvor.; in that position, it would have been deleted as a gloss. A change of punctuation sets everything right: ... Kai lri ivor.; "topal vpwv aKoVual. 7tWr.; extl 7tpor.; 't'o eppciJ­ uOal 7tiVtlV J4 yaOwv; Ou"apWr.;, IPaVal, au''' au't'or.; eppwpal. "I still require to hear from one of you. How is Agathon dis­ posed with regard to engaging in drink?" "I'm not up to it either," he said. The emphatic evor.; ••• vp,wv refers to Agathon, who is elegantly addressed in the third person. He understands Eryxi­ machus' meaning and answers accordingly. A passage from Bos­ well's Life of Johnson will illustrate the usage: RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 R.RENEHAN As Mr. Burke and Mr. Langton were walking home, Mr. Burke observed that Johnson had been very great that night; Mr. Langton joined in this, but added, he could have wished to hear more from another person; (plainly intimating that he meant Mr. Burke). "0, no (said Mr. Burke) it is enough for me to have rung the bell to him."11 377 Plato himself in the Phaedrus (228A-C), availing himself of this same device, has Socrates address Phaedrus to his face in the third person for some dozen lines. VI Symposium 181 D nape(JKeva(Jf.1ivOl yap oif.1,al eialv 0" evrevOev dpx0f.1,eVOI epiiv we; rov Ploy anavra (JvveaOf.1,GVOI Kai KOlVY aVf.1,p,WaO­ f.1,evOl, dAA' OVK eeananlaavree;, ev drppoavvy Aapovree; we; veov, KarayeAuaavree; oixrjae(JOal en' dAAOV dnorpexovree;. "napeaKevaaf.1,eVOI KrA. For the change of construction from we; with fut. partie. to (fut.) infin., cpo Charm. 164D, Rep. 383A nOlelV we; f.1,rjre ... ovrae; ... f.1,rjre ... napo.yelv. The clause ev drppoavvy ... veov is best taken closely with the preceding participle, and KarayeAuaavree; ... dnorpexovree; closely together." Bury ad loco "dAA' OVK .. . 0ixrj(JeaOal is coordinated with epiiv, and both de­ pend on d3 nape(JKevaaf.1,eVOl ... eialV • .•. " Dover ad loc. This latter explanation can hardly be correct; the natural contrast is be­ tween we; rov Ploy ... aVf.1,p,WaOf.1,eVOI and eeanarrjaavree; •.. dno­ rpexovree;, which express, respectively, a noble and ignoble inten­ tion. The main statement runs from napeaKevaaf.1,eVOI to epiiv; the rest of the sentence is subordinated to these words. Moreover, the shift of tense from the present epav to the future oixrjaeaOal (on which Dover is silent) is very harsh, if the two infinitives are to be coordinated. (Herwerden conjectured oTxeaOaz, a Procrustean so­ lution.) Furthermore napaaKeVo.(eaOaz with the future infinitive is a doubtful construction. (Still, it may occur at Xen. Cyr. 7.5.12, and I would be prepared to admit it here, were there not other ob­ jections to this interpretation.) Bury's interpretation is closer to 11 Boswell's Life of Johnson . .. ed. G. B. Hill, rev. L. F. Powell, IV (Oxford 1934) 26-27. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 378 SOME PASSAGES IN PLATO the mark. He rightly sees that the basic contrast is between w<; ... UVIlP,WUOjJ£VOl and OiXr,UBU(}az, and his analysis of the relationship to one another of the various cola, from e~ana!1juavrB<; to ano­ rpixovrB<;, is correct. However, he does not really explain the shift in construction from w<; with future participle to simple future infinitive, nor are his parallels particularly close. The best explana­ tion is to assume, as so often in Plato, an anacoluthon, or rather a constructio ad sensum. w<; with the future participle is used to indicate purpose or intention; another common way of expressing intent is by lliAAw with the infinitive. Plato wrote Oixr,uBu(}al as if he had begun aAA' OU lliAAovrB<; ... ; the shift was all the easier since Plato had not repeated w<; in the second section (i.e., he did not write aAA' OUX w<;). The suggestion that oixr,uBu(}al-here clearly expressing intention-is used specifically on the analogy of IliAAw c. inf. may seem too precise; but compare W. W. Goodwin: "The future infinitive with IliAAw forms the only regular excep­ tion to the general principle which restricts the use of the future infinitive to indirect discourse" (Syntax 2 §7S). It remains to discover Plato's motive for the change to the infini­ tive, which seems to have been quite deliberate, and not the result of careless composition. In the two contrasting clauses (w<; rov Piov ... UVIlP,WUOIlBVO, - OUK e~anarr,uavrB<; ... anorpixovrB<;) the key concepts are UVVBuollBVOI Kai KOIVU UVIlP,wuoIlBVOI and oiXtl­ uBu(}al, referring respectively to the faithful and fickle lover. In this sentence there are no less than eight participles, four of them occurring in the last section (beginning with aAA' OUK). Had Plato written the participle oiXT/UoIlBVOI in the midst of these other parti­ ciples, it would have lacked the emphatic prominence which the infinitive succeeds in bringing out. There was another reason for avoiding the participle here. oiXBu(}al is frequently used with a supplementary participle (oiXOllal anuiJv etc.). Here oiXtlue(J(}al dnorpixovre<; is such a construction. One will search a long time to find an example of the participle of oixollal so collocated with a supplementary participle. A combination such as oiX1JuOIlBVOI anorpixovre<; simply does not sound right, and Plato instinctively avoided it. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 Symposium 182A-B R. RENEHAN VII Kai ~~ Kai () nepi rov epwra VO/lOe; tv /lev raie; aAAale; nOAeal voijaaz pfj,blOe;, anAWe; yap Wplaral" 6 <>' iv(Jci<5e Kai • A 1:' ", ''H' I: ' , "B ev aKeOalflOVl nOlKIAOr;. ev Alol flev yap Kal ev 010)- roie;, Kai OD /liT aO(poi Aeyelv, anAme; vevo/lo(Jer1'/raz ... rije; <>i lwviae; Kai aAAo(J1 nOAAaxov aiaxpov VeVO/llaraz, aaOl uno pappcipOle; oiKovalv. rije; be 1wviae;: !Ole; be 1wviae; ci. Ast: r6 ~e 1wvi", Thiersch 379 "r~e; ~e lwviae;. The genitive is taken by Hug as dependent on nOAAaxov, by Stallb. as dependent on oaOl, 'vel potius ex demon­ strativo ante oaOl intelligendo'. Hug quotes Xen. Hell. IVA.16 nOAAaxoae Kai rije; :4pKa~iae; i/lpaAOvree;." Bury ad loc. "rije; ~i ... nOAAaxov 'in many parts of Ionia and in (sc. many) other places'." Dover ad loc., who, to judge from this translation, agrees with Hug in making lwviae; dependent upon noAAaxov. Stallbaum's proposal to govern 1wviae; by aaOl is unnatural and, indeed, desperate. Nor can lwviae; be governed by nOAAaxov (as Hug, Dover, and others); it should have been obvious that rije; lwviae; and aAAo(Jl nOA),axov are coordinate phrases linked to­ gether by Kai. In Xenophon Hell. 4.4.16, compared by Hug, the word-order makes all the difference; the two passages are not at all parallel. For true parallels see Protagoras 326DE Kai nap' v/liv Kai aAAo(Jl nOAAaxov and Republic 394c ev r:e rv rwv enwv nOl1jael, nOAAaxov be Kai aAAo(J" which show both that the two phrases are to be taken as contrasting coordinates and that aAAo(Jl nOAAaxov is a set phrase to be taken by itself. (See also Symp. 209E, Resp. 440A OVKOVV Kai aAA081, erp1'/v, nOAAaxov ala8avo/le8a, Xen. Cyr. 7.1.30 nOAAaxov /lev ouv Kai aAAo(J, ... Kai iv rovup be . ... ) Two words have caused all the difficulty, lwviae; and oaOl: what is the construction of the genitive and what is the antecedent of aaOl? Both uses are in fact quite idiomatic. oaOl is a constructio ad sensum, introduced for variety where ou might have been ex­ pected; it is as if ev "/WalV Kai aAAOle; nOAAOie; had preceded; compare, immediately above: (1) tv "HAl<>1 ... Kai (2) iv BOlwroic; Kai (3) OD /l~ aorpoi Aeyelv. The other stumbling-block, rile; ~i '/wviae;, is not 'governed by' any word in the sentence; it is rather one more example of the common genitive of connection, placed (with connective, but with or without preposition) at the beginning RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 380 SOME PASSAGES IN PLATO of a new sentence as a separate colon: "But as regards Ionia, and in many other places .... " For the usage in general, see Fraenkel on Aeschylus' Agamemnon 950, with references. The same construction occurs at Phaedrus 247 A: JJivel yap E(Jria tv (}ewv OrKqJ fJ.OVl'r nov i5i a"t"tcov OUOl tv reP rwv &hi5eKa apl(}fJ.eP reraYfJ.eVOI (}eoi apxovrer; Ilyovvral Kani rd~lv 'Iv eKaurOr; trdx(}". rwv ~e dAAWV O(JOl is not to be rendered "But as many of the others as ... ," as if rwv dAAWV were genitive dependent upon oaoz. Hackforth, for example, has understood the sense: " ... but for the rest, all such as are ranked in the number of the twelve .... " rwv ~e dAAWV is a separate genitive of connection; editors should have printed a comma after it. So also in Symposium 221c dAAcl rwv f.leV dAAWV htlr'1~evf.larwv, rax' dv ru; Kai nepi dAAOV rOlavra elnOI (where editors omit the comma after htlr'1~eVf.larwv). Com­ pare Phaedrus 250c: nepi ~e KaAAom;, wanep elnof.lev, f.l8r' eKeivwv re e).af.lnev ov KrA. That nepi ~e KaAAov~ was felt to be a distinct colon is shown by the fact that the subject of the main verb eAap,­ nev is also K(iAAO~ . VIII Symposium 184D-E orav yap ei~ ro auro eA(}W(JlV 8pa(Jrrj~ re Kai na'~'Ka, VOf.lOV exwv t.Karepo~ ... 0 f.lev ~vva.f.leVO~ ei~ rpPOV'1(JIV Kai '!lTV aAA'1v dperr,v aVf.lpdAAe(J(}al, 0 i5e ~eof.levo~ el~ naii5evalv Kai rr,v aAA'1V aorpiav Kraa(}al, rore i51j ... avf.l­ nfnrel ro KaAOv eivaz nal~'Kcl epaar6 xapfaa(J(}al, dAAO(}, t5e ovt5afJ.ov. elr; post &ofJ.eVOr; del. Schutz Kriiu(}al: furau(}al Schanz: Krau(}ai rz Hug The soundness of the MSS here has often been questioned. I quote Bury's note as typical: "el~ naf~ev(Jlv ... Krii(J(}al. If the text is right we must suppose that Kraa(}al is here equiv. to w(Jre Kra(J(}al, appended to the main verb eVf.lpa)'Ae(J(}al which is to be supplied with el~ naft5ev(Jlv K!A. (so Vahlen). Of the corrections suggested ... Schanz's is the neatest, but spoils the sense-balance with ~Vf.lpaAAea(}al. The corruption is, perhaps, to be sought else- RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 R. RENEHAN 381 where: the expression rqv dJ..J.."v aorpiav is open to suspicion, since aorpiav as here used after dJ..J..1'fv stands as a generic sub st. whereas aorpia has just been termed (184c) /-tipoc; aper-qc;: moreover, we should expect that aorpia should itself constitute the Kr-qf-la of the recipient, just as rppOV1'falC; is itself the contribution of 0 f,vf-lpa).­ ).of.levo~. On these grounds, I venture to suggest that another fern. subst., such as ou5axrjv, may have fallen out after 11J..J..1'fv (eK1Cai­ <5evalv for elC; n. is just possible)." Dover too, in his recent edition, follows earlier scholars in expressing doubts about the soundness of the text: "dc; nai<5evalv ... Kraa(}al: if this is what Plato wrote, we must understand rppov1'fazv Kai aperrjv as object of Kraa(}al; but if dc; were deleted (as by Schutz) 1Cai<5evazv ... aorpiav would be the object, and the sentence would be easier to follow." All this is much ado about nothing. The speech of Pausanias, from which this sentence comes, contains a generous portion of those tricks of rhetoric associated in particular with the soph­ ists. llavaaviov <5i navaaf-livov, says the narrator at its conclusion (185c), <>1<5aaKOVaz yap f-l8 iaa ).iyelv ovrwai 0" aorpoi, and the present sentence is clearly an instance of ro iaa J..iYelv, wherein rhetorical balance counts for more than normal idiom: (i) 6 J1iv l5vvdJ18voc;-6 <5e &6J18voc; (ii) dc; rppov11rJlvKai rqv cUJ..11v dpe!1jv­ dc; nai&valV Kai rqv dJ..J..11v aorpiav (iii) aVJlPdJ..J..ea{)al-Kraa{)al Kraa()az is in obvious sense responsion to aVf-lpaJ..J..ea()al; note the force of the present, 'acquire' (not KeKr-qa(}az, 'possess'). The one makes a contribution; the other acquires the same. Render the disputed phrase quite literally: " ... the other needing to make an acquisition in regard to nai<5evau; and the rest of aorpia • •.• " No object need be supplied with Kraa()al; the verb is used absolutely.12 Objections to the transmitted text clearly have taken their start from the assumption that dc; 1Cai<5evazv ... Kraa(}al is awkward, or even impossible, Greek. I have argued that a desire for artificial balance determined the choice of phrase; no one would deny that Kraa()al followed by a direct accusative object would have been 12 Even were it necessary to supply an object, the comments of Bury and Dover would still be misleading. For here rppOV'fUlr; Kai ~ tUA." dpe,~ and 7tai(jevUlr; Kai ~ tUA." uorpia are, for all practical purposes, synonymous; the change of diction was determined solely by a desire for stylistic variation, and it is a mistake to press any difference of meaning. This is apparent from the context: 0 r5e6pevor; Kniu(}al is here equivalent to 'the one needing to receive a contribution' and corresponds to 0 i5vvapevor; uvppaMeu(}al. Obviously, giver and recipient are concerned with the same contribution. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 382 SOME PASSAGES IN PLATO more usual. Possibly the expression is unusual; certainly it is Pla­ tonic. For this use of de; (= 'in regard to', 'with a view to'), where another construction might be expected, is a feature of Plato's style. From this same speech compare 184B evepyerovf.J£voe; de; xpr,para (the dative, vel sim., would be 'normal'). So also below in the Symposium, 219 D .•• av(}pw1up rolOvup oi'cp eyw OUK uv QJP1'/V nor' evrvxeiv de; rpPOV1'/(IlV Kai de; Kaprepiav. For further Platonic examples see J. Riddell, A Digest of Platonic Idioms § 115. IX Phaedrus 233n trl oe el Xpr, rOle; oeopiVOle; pUAlara xapi(ea(}al, npoar,Kel Kai rOle; aAAOle; pr, rove; peA riarove; aAAa rove; anopwni­ rove; ev nOlelv· peyiarwv yap anaAAayivree; KaKWv nAeiar1'/V XUPIV avrole; eiaovraz. Kai rOle; aAAOle; has bothered many; consequently Kai rwv aA­ AWV was printed in the Aldine edition and Badham proposed Kdv rolc; aAAOle;. De Vries ad IDe. defends the MSS: "The readings Kai rwv aAAWV (Aldina, Heind., St., Sch.) and Kdv rOle; aAAOZe; (Bad­ ham, Vollgr., Buchw., accepted by Hackf.) are evident attempts to make the text smooth. Ficinus (Bekker) already rendered rOle; aAAOle; well by 'omnino'. For the dativus limitationis which he rightly sees in rOle; ClAAOle;, Verd. 271 refers to 234e7 [ra ... aAAa, accusative!], Lysis 21Se apu ye OAq1 nvi ef,anarwpe(}a; Rep. 430A enazoevopev povazKY Kai yvpvaanKY, Hom. Od. 18.234, Hdt. I 29, Thuc. IV 73, 4, K.G. I 437 f. (Rob., too, took it as such, as may be seen from his punctuation Kai, rOle; aAAOle;, and his trans­ lation 'par ailleurs'; Rob. PI., however, wrongly takes rOle; aAAOle; as masculine, 'pour les autres aussi'; Mor. unnecessarily suggests Kai rove; aAAOVe;)." Thus the majority opinion seems to be that Kai rOle; (lAAOle; is eiher corrupt or neuter (or some combination thereof). It is sound and masculine, as Robin once took it-only to change his mind later. For neuter rOle; aAAOle; = omnino no one has produced a true parallel; I doubt that any exists. Neither the accusative ra aAAa (common) nor the dativus limitationis of other substantives is pertinent, much less decisive. What is wanted is another instance of rOle; aAAOle; so used; that usage, if it exists, is unknown to me. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 R.RENEHAN 383 (On this compare below.) To understand the sentence, construe as follows: ... ei Xp~ [sc. rove; epWj.1BvOVe;] roie; JeOj.1iVOle; j.1oJ.lara xapi(ea(Jal, npoa~Kel Kai roie; (fJ .. AOle;-etiam ceteris hominibus praeter rove; epwj.1ivove;- ... rove; dnopwrarove; eV nOlelv' j.1eyiarwv yap dna,V.ayivree; KaKwv nAdar'lv xaplv auroie; [sc. rolc; (iV.Ole;] ei'aovraz. Note that aMofc; has no reference unless rOl<; 0).}.0l<; be both sound and masculine. The argument-which comes from the oratio erotica which Plato has fathered on Lysias-is a reductio ad absurdum. The general thesis of this speech is that a boy should bestow his favors upon the non-lover rather than the lover. The particular argu­ ment here is: "If a boy should yield to those most in need [i.e., to lovers], it follows that it is appropriate for 'the others' also to treat well specifically those who are most at a loss [in any human sphere] .... " The consequences of such a position are then illus­ trated; for instance, consistency would require that one invite to a private dinner not his friends, but beggars and those in need of a meal (233DE). That such is the correct interpretation and that ro~ (i).AOle; is sound is proved by the words roie; (iAJ.OlC; themselves. For it does not appear to have been remarked that in this speech 01' (1)),01, 'the others', is frequently used, almost in a technical sense, in explicit contrast to epwVTSe;. The relevant passages are self-evident and need only be set forth, not interpreted: (i) erOl/loi d(Jl [sc. ol epwvre~] ... roi~ (j)J .. Ol~ aneXBaVO/leVOI roi~ epwJ1l5,vol~ Xapi(e(JBQl (231c) (ii) d /lev eK TWV epwvTwv TOV po..n(Jrov alpoio, ee; oAiywv dv (J0l ~ eKAee;l~ efr/, d J' eK TWV d),AWV (231D) (iii) dKOe; e(Jn rove; /lev epwVTa~, OUTwe; av oio/levove; Kai uno TWV dAAWV (1]Aou(JBQl w(Jnep avrov~ vrp' aVTwv (231E-232A) (iv) Jlonep Kai Tae; npoe; rove; dAAOVe; TWV epw/lf.vwv (Jvvov(Jiae; anoTpenov(Jlv [sc. of epWVTe~] (232c) (v) Tomura yap 0 epwe; em&iKVVral' JV(JTvxovvrac; /lev, Ii /lr, AVn1]V roie; d).AOle; napexel, aVlapa nOlei VO/li(eIV (233B) (vi) orJJe of J,anpae;af.J.8vO, [sc. ol epwvree;] npoe; TOVe; dAAOVe; rpIAOTl/l"(JOVTaI (234A) 13 I return for a moment to the suggestion that roie; UAAOle; in 233D is a neuter = omnino. The reason why roie; UAAOle; is not 13 Note also 252n (from the great myth), npor; Te TOUr; epw;dvovr; Kai rour; d22ovr; of.l12ei Te Kai 1tpourpeperal. We are perhaps justified in regarding this as a mannerism of Plato's and one more, albeit minor, argument for the Platonic authorship of the speech attributed to Lysias. RENEHAN, R., Some Passages in Plato , Greek, Roman and Byzantine Studies, 22:4 (1981:Winter) p.371 384 SOME PASSAGES IN PLATO found so used (quod sciam) is that in the dative ending -Ol~ neuter and masculine are not distinguished; the gender is unmarked and ro;~ aAAol~ = omnino would be ambiguous. 14 Hence the prefer­ ence for ra clAAa everywhere in ancient Greek. Consider Phaedrus 234 c: ri aOl rpaiveral, W EWKpare~, () AO'YO~; OOX v1teprpvw~ ra re clAAa Kai ro;~ OVOJ.laalV eipija()al; Here the accusative of respect ra clAAa is used despite the fact that it is collocated with a dative of respect, ro;~ ovoJ.laGlv. In other words, Plato himself in this very dialogue avoids ro;~ aAAOl~ where one might most expect it. There is no doubt of the soundness of the text; Plato repeats the phrase verbatim below, 257 A. Elsewhere15 I have argued for the reality of the collocation of accusatives and datives of respect in Greek; Phaedrus 2340 and 257 A provide the clearest evidence to date for the existence of the usage in classical prose.16 THE UNIVERSITY OF CALIFORNIA AT SANTA BARBARA June, 1981 14 One might wish to raise the same objection against a masculine Toi, dUo" in 2330. There is far less reason to do so. In the sequence 1lpoO?iKe,! dative! infinitive an ancient Greek was not likely to be confused, given the frequency of the construction 1lpoO?iKe, cum dat. et infin. Context determined the sense. But even if the passage is felt to be ambiguous, the fact is that 1lpoU11Ke,1 masc. dat.! info is a normal and frequent collocation, whereas Toi, dA.M" = omnino is apparently unattested. 15 Greek Textual Criticism (Cambridge [Mass.] 1969) 109-12; CP 75 (1980) 245-46. 16 I am grateful to my students Lorna Holmes and Morgan de Tarr for reading a draft of this paper and making a number of helpful suggestions.