Notes on Donatus' Commentary on "Adelphoe" Grant, John N Greek, Roman and Byzantine Studies; Summer 1971; 12, 2; ProQuest pg. 197 Notes on Donatus' Commentary on Adelphoe John N. Grant THE ANCIENT COMMENTARY on Terence that has come down to us in the name of Aelius Donatus gives interesting information about the Greek models and the changes which Terence made in his adaptations. The commentary, however, is not the original one of Donatus but a later syncretism of part, at least, of Donatus' work and scholia drawn from other sources.! There are scholia which are misplaced or which contradict each other on the same point and not a few which are unworthy of any grammarian. Difficulties are increased by the fact that the text is often corrupt, particularly when Greek quotations are cited. Thus many of the problems con­ cerned with the relationship of Terence to his Greek models remain unsolved. Yet full use of the evidence of the commentary has not been made. I hope to show here that if the wider context in which the relevant scholia stand is analyzed, the relationship of Terence's Adelphoe to the Menandrian original can be clarified with respect to the words of Micio at lines 43-44 and to the same character's opposition to a marriage with Sostrata at the end of the play (lines 934ff). I also suggest that the statement in a scholion on line 275 that in Menander Ctesipho intended to commit suicide is erroneous. In addition an emendation of the scholion on line 323 is offered. I 43 ... 2 ET QVOD FORTVNATVM 1ST! PVTANT Romani scilicet, qui caelibem quasi caelitem dicunt, et item Graeci, apud quos sunt huiusmodi sententiae-IIal_u!nAoc €XOL tVL£vV 7TWAOVJ.LEVOVt et alibi yaJ.LEL IIaw/JtAoC' yaJ.LELTW. Kat yap ~SLK'YJaE J.L€. 3 FORTVNATVM 1ST! PVTANT utique uxorem non ducere. dicit autem Romanis id 1 See Aeli Donati Commentum Terenti, ed. P. Wessner (Leipzig 1902-08) I praef xlv-xlvii. In this article the text of Donatus is cited from Wessner's edition. 197 198 NOTES ON DONATUS' COMMENTARY ON ADELPHOE videri, quos spectatores habet. Menander tTwfLaVKapta7]fLvt yvvaiK' OU AafLf3avw. 4 ET QVOD FORTVNATUM 1ST! PVTANT quidam putant sic pronuntiandum < et quod Jortunatum isti putant uxorem' et haec bona et concinnata locutio est. It is unfortunate that the first part of the quotation from Menander in scholion 3 is corrupt, since one would like to know whether Terence diverged from his Greek model at this point. The corruption is such, however, that none of the proposed emendations has won overwhelm­ ing support.2 The suggestions are often equally plausible-or im­ plausible, if you will-palaeographically, and in the final analysis must all have depended on whether or not their authors thought that Terence followed his original closely at this point. Obviously such emendations cannot by themselves provide an answer to the question of Terence's fidelity of translation at lines 43 and 44 of the Adelphoe. An additional criterion has to be found. 3 The attempt will be made here to show that if the relevant section of the Donatus commentary is examined more closely than has been done in the past, we do in fact have a yardstick other than palaeographical plausibility by which to judge those conjectures that have been made. First of all it is necessary to look at the three scholia cited above both in terms of content and in the arrangement of that content. They have been prompted by two questions. To whom does isti refer? Should one take the relative clause quod Jortunatum isti putant to refer to uxorem numquam habui or to uxorem alone? As far as the second question is concerned, the scholiast favored (correctly, I believe) taking uxorem numquam habui as the antecedent. Otherwise, he would hardly have troubled to write scholion 2, which is irrelevant if uxorem were the antecedent. Moreover, the manner in which the alternative inter­ pretation is given does not indicate the writer's approval of it. 4 One 2 In the discussion of this verse in the Teubner edition of Menander (A. Korte/A. Thier­ felder, Menandri quae supersunt: Pars altera [Leipzig 1959] 16) the conclusion versus ncndum certo sanatus videtur is a judicious comment on the attempts that have been made to emend this line. 3 Cf the statement of P. Sipkema, Quaestiones Terentianae (Amsterdam 1901) 64: sed me iudice versus Menandrei condicio ea est ut nihil nobis prosit in discernendo utrum presse secutus sit graeca Terentius annon. 4 This is, of course, based on the assumption that all the scholia were written by the same person. It is possible that the whole of scholion 4 was written later than the other scholia, but I see no reason to doubt that utique uxorem non ducere of scholion 3 and quidam putant sic pronuntiandum 'et quod fortHnatHm isti putant uxorem' of scholion 4 were written by the same JOHN N. GRANT 199 wonders therefore whether the verse quoted from the Menandrian play was a decisive factor in influencing the commentator in his choice of antecedent for quod Jortunatum isti putant. Was it cited as conclusive evidence for the interpretation he preferred? If, however, one looks more closely at the scholion in which the Menandrian line appears, one sees that the first half (utique uxorem non ducere) is related to the problem of punctuation, while the remainder (dicit autem Romanis id videri, quos spectatores habet) seems to be a continuation of the topic of scholion 2 and to affirm that, although among Romans and Greeks there is evidence of a common attitude to marriage, isti must refer to the Romans. Thus the first half of scholion 3 is closely connected with scholion 4, while the second half of scholion 3 is closely connected with scholion 2. I suggest, therefore, that when the ancient commentary took the form in which it has come down to us the first half of scholion 3 was erroneously placed between parts of the same scholion.5 It has now to be decided whether the Menandrian verse should still follow dicit autem Romanis id videri, quos spectatores habet or whether it goes with utique uxorem non ducere and was quoted baldly and with­ out connective as support for the punctuation which the comment utique uxorem non ducere requires. But in other places in the com­ mentary where an alternative interpretation is reported and we are given information of the Greek model, the reference to the Greek play appears after the former: thus An. 483.3, FAC ISTA VT LAVET 'ista' quae ex puerperis sordebant. qUidam 'ista' ips am puerperam dicunt-sic enim et Menander AOVW'-' (XJh~v alhtKa-sed ... ; An. 592.1, QVIDNAM AVDIO legitur et 'audiam' ; Menander enim sic ait ,-t 7TO'-' aKoVCOfLat; Eun. 46.7, NON EAM NVNC QVIDEM 'non eam' Probus distinguit; iungunt qui secundum Menandri exemplum legunt. Thus, if we take the Menandrian verse with utique uxorem non ducere and link both parts with scholion 4, we find a divergence from the usual pattern, unless we assume (1) that no part of scholion 4 was written by the author of scholion 3 and that scholion 4 was a later addition, or (2) that the Menandrian hand. Since that individual preferred to take uxorem numquam habui as the antecedent, the final part of scholion 4 (et haec bona et concinnata locutio est) may be a case of' damning with faint praise' or the comment of a subsequent writer who found the alternative interpreta­ tion attractive. 5 Misplaced scholia are not infrequent in the commentary. See R. Sabbadini, "Gli scolii donatiani," Satal 3 (1895) 340-41, for examples in the Eunuchus. 200 NOTES ON DONATUS' COMMENTARY ON ADELPHOE scholion was also (i.e. in addition to utique uxorem non ducere) misplaced. On the other hand, if the Menandrian line is taken with dicit autem Romanis id videri, quos spectatores habet as the concluding part of scholion 2, the structure of scholion 2 would resemble that of the scholia on An. 483, 592 and Bun. 46. The evidence suggests, therefore, that the Menandrian line should be taken with scholion 2, and that it was cited in connection with taking isti to refer to the Romans. In addition to the scholia cited above, other testimony of the diffi­ culty which isti caused in antiquity is provided by the final part of scholion 5 on this line in the Donatus commentary ('isti' autem id est hi qui a me dissentiunt) and by the presence of a scholion in the Bem­ binus (isti: qui uxores habent). My own view on the demonstrative is that Micio is referring, strictly within the context of the play, to his fellow citizens. and that isti indicates more forcefully than, for example. quidam that he himself does not share their belief. Donatus, however, states that isti refers to the Roman audience. The ancient commenta­ tor was echoed by Karl Dziatzko and Robert Kauer,6 who followed Joseph Bach' in believing that there is inherent second personal force in the demonstrative iste. But Bach's thesis was ably challenged by Ruth Mildred Keller, who showed convincingly that the primary force of iste was strongly deictic.8 Admittedly, one explanation of Donatus' denial of the possibility of taking isti to refer to the Greeks would be that he recognized a second personal force in iste. His pur­ pose in citing the Menandrian line would then be uncertain. It might have been quoted to indicate that there was an allusion to the au­ dience in the Greek play also. But it may have been adduced to show that there was no reference to the audience in Menander and that Terence differed from his original at this point. Other notes on this demonstrative in the ancient commentary show, however, that iste was thought to have strong deictic or derogatory force.9 It seems, therefore, that the Menandrian line was cited because it was on the basis of this that the scholiast stated that isti referred to the Romans. Accordingly, some conjectures about the content of the Greek verse can be made. The Menandrian Micio did not, for example, directly 8 K. Dziatzko/R. Kauer, Terentius: Adelphoe2 (Leipzig 1903) ad loco 7 De usu pronominum demonstrativorum apud priscos scriptores Latinos, in W. Studemund (ed.), Studien auf dem Gebiete des archaischen Lateins II (Berlin 1891) 145-415. 8 "Iste Deiktikon in the Early Roman Dramatists," TAPA 76 (1946) 261ff. 9 See for example the notes in the commentary on An. 15, Ad. 377, 388, 981. JOHN N. GRANT 201 address the audience in the second person (=quod Jortunatum putatis) , since a change from a direct address to quod Jortunatum isti putant would provide support for taking isti to refer to the Greeks. Secondly, it is most unlikely that quod Jortunatum isti putant corresponds exactly to what was said at this point in Menander, since an exact translation would not help to show that isti must refer exclusively to the Romans. The conclusion that one must come to is that in Menander the old man made no comment on the opinion held either by the audience or by a third party but stated his own belief that bachelorhood was a blissful state. Thus the fragment as printed in the Teubner edition­ XW JLuKapu5v cf>U, YVVUtK' Oll AUJLf3avw1o-is less attractive in sense than a suggestion made by Louis Havet-To JLUKapLOV TO 7Tavv, YVVUtK' Oll AUfLf3&VW.11 Finally, the change which Terence has made at this point may be connected with Demea's proposal at the end of the play that Micio should marry Sostrata (929ft} Terence may have tried to soften the shock of Micio's acceptance by changing what was a personal opinion of the old man on the blessings of being a bachelor in the Greek play into a comment, implying disagreement, on the attitude of others towards marriage.12 II 275 2 " ''/'' ~, • •. PAENE EX PATRIA U7TOCLW7TTJCLC EV'f'TJJLLCJLOV Xupw. Menander mori illum voluisse jingit, Terentius profugere. Different reasons have been put forward for the change which Donatus tells us Terence made here. It has been suggested that it was made out of concern for consistency in the characterization of Ctesipho,13 in order to tone down the emotion,14 or in the interests of 10 Based on a suggestion of A. Saekel, Quaestiones comicae de Terenti exemplaribus Graecis (Berlin 1914) 54, although the punctuation is different. 11 "Un fragment de Menandre, Adelphes," RevPhil45 (1921) 86-87. Suggestions similar in sense to Havet's were proposed by F. Nencini, De Terentio eiusque fontibus (Liburni 1891) 135, and Dziatzko/Kauer, op.cit. (supra n.6) ad loc., but these are less attractive palaeo­ graphically or metrically. 12 W. G. Arnott, "The End of Terence's Adelphoe: Postscript," G&R 10 (1963) 143, rightly emphasizes that Micio does not concur with the opinion of isti. 13 Dziatzko/Kauer, op.cit. (supra n.6) 16. 14 G. Jachmann, RE 5 A (1934) 615. 202 NOTES ON DONATUS' COMMENTARY ON ADELPHOE realism. IS I fail to see how the change .. dient der Dampfung des Affekts," since it is not as if Ctesipho is himself speaking and contem­ plating what he should do in light of the apparent loss of his girlfriend. As for the consistency of characterization, if one wished to go into the psychological makeup of Ctesipho, a decision to commit suicide could be regarded as more consistent with his portrayal as "ein haltloser und unselbstandiger ]iingling" than a decision to leave his homeland. The other motive for the change, a concern for realism on the part of Terence, would be more readily acceptable if the contemplation of suicide was not such a common motif in New Comedy16 and it was not hinted at in other plays of TerenceP Nevertheless, this concern for realism still seems to be the best explanation for the change if the statement in the scholion is accepted. I wish to offer here the suggestion that, while there was a difference between the Terentian and Menandrian plays at this point, the differ­ ence was not what the Donatus scholion tells us. Another solution for young men in Ctesipho's plight was to go overseas, often to fight in the Hellenistic armies.1s An interesting example of this can be found in the Samia of Menander. In this play Demeas had suspected that his adopted son Moschion was the father of a child by the Samian woman. The young man reveals the facts of the matter to Demeas, but Moschion's anger at the mistrust shown by his father breaks out in a later monologue (281ff) where he states that were it not for other factors he would not wait around to be accused a second time: '''' , '1" " OVK av 7TapOV'Ta y aV'TtC 'rJ'T£aaaTO , , A' ,~, >\ \" J..8 \ aVTOV JLE TOtOVT OVOEV, a/\/\ a7To'fJ apEtC , A '\ "" ~ \ 'B ' EK T'rJC 7TOI\EWC av EK7TOuWV EtC aKTpa 7Tot .,. TT , ~'f3 ' 'Y 'A 'rJ napLav ULE'rpL OV aLXJLa~wv EKEt. The appearance of the verb &'7Toc/>8ELpEC8at in this context in the sense 'to be gone' is interesting.I9 Is it possible that this verb appeared in 15 H. Haffter, "Terenz und seine kiinstlerische Eigenart," MusHelv 10 (1953) 96; O. Rieth, Die Kunst Menanders in den 'Adelphen' des Teren{. Mit einem Nachwort herausgegeben von Konrad Gaiser (Hildesheim 1964) 40. 16 See for example Men. Pk. 255, Plaut. Asin. 606ff, Cas. 307, Cist. 248, Mere. 472, Pseudo 88ff. 17 An. 322, Hec. 282, Phorm. 686 (see K. Dziatzko/E. Hauler, Terentius: Phormio4 [Leipzig 1913] ad loc.). 18 See for example Plaut. Mere. 644ff, Ter. Haut. 117. 19 Cf Men. Sam. 158, Eur. HF 1290. JOHN N. GRANT 203 'A8EAcf>OL fY in the verse which was the model of lines 274-75 in Terence's play (tam ob parvolam I rem paene e patria-turpe dietu!), and that the information about a difference between Terence and Menan­ der with respect to Ctesipho's intentions rests on a literal interpreta­ tion of this Greek verb? Let us suppose that there was such a misunderstanding. Is there any indication of what prompted it? Originally the Greek verse, or at least some part of the verb a,7T0cf>0ElpEcOaL, could have been quoted to show that there was no aposiopesis in Menander, but it may have been cited as a guide in deciding which verb would have completed the construction in Latin. The first scholion in line 275 reads PAENE EX P ATRIA deest 'fugere', quia amatores eomici cito eomminantur patriam se deserturos, ut amieam eonsequantur. But because turpe dietu suggests a stronger term than fugere and because of the alliteration in paene e patria, it is more likely that we are to think that the verb which Aeschinus shrinks from saying is perire (ef Plaut. Capt. 537, periisti e patria tua). Now the fourth scholion on line 274 reads et deest 'perire', sed r0 EVcf>YJfLLCfL0 taeetur. I would conjecture that because of the similar usage of perire and a,7TOcf>OElpEcOaL the Menandrian quotation was originally part of scholion 274.4, that the two parts of the scholion became separated and that a late scholiast (or the compiler of the commentary?), confronted with the Menandrian tag in isolation, jumped to the hasty conclusion that in Menander the young man had contemplated suicide.20 In support of this reconstruction of how such a misunderstanding could have happened, it may be said that the same individual could not have written both 274.4 and 275.1. The former is superior in its choice of verb and more likely to go back to the original notes of Donatus. Moreover, this scholion (274.4) is clearly misplaced. It follows a scholion written under the lemma TAM DB PARVVLAM and has no appropriate lemma of its own. One concluding point. Later in the play (384-85) Demea prophesies what he sees in store for Aeschinus: videre videor iam diem illum quom hine egens profugiet aliquo militatum. 20 Traces in the Terentian play of the suicide have been seen in nunc vivo (261), if taken literally: so Rieth, op.cit. (supra n.15) 41. But it is equally possible that the presence in the commentary of a scholion on nunc vivo might have contributed to the misunderstanding of the Greek verb. 204 NOTES ON DONATUS' COMMENTARY ON ADELPHOE How much neater would the irony be if in the Menandrian play Ctesipho, the son whom Demea believed rei dare operam, ruri esse parcum ac sobrium (95), was himself on the point of going overseas to take up arms before the timely intervention of Aeschinus! III 323 QVID FESTINAS MI GETA Probus personae assignat hoc Sostratae, Asper non vult ad omnia servum respondere, sed nutricem putat hoc loqui. servum] Sostratam Westerhovius, eram Smutny, utrumque male; nam ad omnia Geta responderet, si Sostrata totum hoc QVID FESTINAS ••. RECIPE diceret The commentary records here a difference of opinion in antiquity as to the speaker of the words qUid festinas, mi Geta? an imam recipe (323-24). The Terentian manuscripts follow Probus in giving them to Sostrata, but I am inclined to follow Aemilius Asper and some modern editors (Dziatzko, Stampini, Fleckeisen2, Kauer-Lindsay) in assigning them to Canthara. The double response or reaction would be quite in keeping with Terence's technique,21 and elsewhere in this scene there is a tendency to rob Canthara of the small part that she has (see the assigning of parts in the manuscripts at lines 336 and 343). The problem which I wish to touch on here, however, concerns the text of the scholion in the Donatus commentary. Wessner's criticism of the emendations of Westerhovius and Smutny is justified, but the text as he left it makes little sense. The second part of the scholion (Asper non vult ad omnia servum respondere, sed nutricem putat hoc loqui) would be comprehensible if the problem was whether the words should be assigned to the servus Geta or to the nutrix Canthara, but obviously the choice lies between Canthara and her mistress Sostrata. That is why Westerhovius and Smutny re­ placed servum by Sostratam and eram respectively. Surgery, however, has been applied to the wrong place. I suggest that one should read Asper non vult ad avYjc cf>{AOC 7J 1TAOihoc acf>av~c, ov cV KaTopv;ac EXEtC. 28 See E. W. Handley. The Dyskolos oflvIelllllldcr (London 1965) ad loe. and on lines 807-]{). 208 NOTES ON DONATUS' COMMENTARY ON ADELPHOE The second reason, Micio's failure to marry, would be invalidated when he agrees to marry Sostrata if there was the prospect of an heir from the union; but this is explicitly excluded by Demea, who humorously adduces as a point in favor of the marriage the fact that Sostrata is past the age of childbearing (931). The old age of Sostrata is further emphasized by Micio at 938-39: ego novos maritus anno demum quinto et sexagensumo flam atque anum decrepitam ducam? If the age of Sostrata was not specified, there is no reason why an Athenian audience should not have seen the marriage as a means of producing an heir, given the early age at which girls married in Athens.29 There would then be complete agreement between the VrroK€lJL€va and the ending of the play, and there would be no loose end hanging from this closely woven plot structure. Is it simply coincidence, then, that Donatus records Terence's variation from his model, not at the beginning of Micio's resistance (e.g. at 934), but at line 938, where Micio's own age and (in 939) Sostrata's age are raised as objections to the marriage? If one looks at the Donatus scholia cited supra, two points can be made. The second scholion on 938 is pointless, uninformative, and unlikely to be the work of Donatus,30 and, once this scholion is expunged, a scholion on 939 comes between two on line 938. It is possible that this inter­ vening scholion (939.1) is misplaced,31 but another explanation which is linked with the remarks made above can be put forward. I suggest that 938.2 was a late addition to the original notes made by Donatus and that this scholion has split two parts of one scholion which originally read like something in this vein: apud Menandrum senex de nuptiis non gravatur "afLf3avw suggests that the Menandrian Micio did not regard himself as being too old for marriage, while the Latin uxorem numquam habui gives the impression of a much older man. The conjecture that Terence was responsible for the mention of Micio's age (and of Sostrata's) was made by Otto Rieth33 on the basis of the differences noted here. But he further be­ lieved that the whole section (934-46) was the work of Terence. I have tried to show, on the basis of (1) the inherent implausibility in be­ lieving that the Menandrian Micio readily agreed to marriage while resisting Demea's other proposals, (2) the agreement of 934-46 with the motif of the reversal of roles at the end of the play, (3) the greater economy of the V7TOKE{fLEva if a child could result from the marriage, (4) the position at which Donatus tells us of the change which Terence made, and (5) the apparent dislocation of scholion 939.1, that Terence's contribution is probably confined to lines 938-39.34 SCARBOROUGH COLLEGE, UNIVERSITY OF TORONTO May, 1971 32 For the order in the second part of the scholion (quotation-adverb-verb), cf Ad. 43.5, et 'putant' bene dicit is cui displicet aliena sententia; 116.2, et 'maximam paTtem' optime dicit; 192.2, et 'quid si nolo' nove dictum. Much more common, however, is the order adverb-verb­ quotation (cf e.g. Ad. 11.3,61.1,65.2,71.2 etc.) or adverb-quotation-verb (cf e.g. Ad. 40.2, 47.1,70.4,89.2, 119.4, 142.1, 143.3). 33 Op.cit. (supra n.15) 119. 34 I believe that Aeschinus was present in this scene in Menander and do not believe that the lie which he utters at 940 (promisi ego iIlis-he has had no opportunity to have made such a promise) is necessarily the work of Terence. But see Rieth, op.cit. (supra n.15) 118.