STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. Classical Philology Volume XI April IOl6 Number 2 THE INTERPRETATION OF ROMAN COMEDY Bt Henry W. Pbescott By way of introduction to various studies of technique in Roman comedy I find it convenient to state briefly, with some illustrative examples, the dominant tendencies, as I see them, in the higher criticism of Plautus and Terence, to suggest the weakness of modern method, and to indicate the possibility of a different point of view and of safer courses of procedure. In such an introduction suggestion rather than demonstration, exposition rather than extended argu- ment, are the limited aims of the paper; and I hope that the brevity desirable in prolegomena will not be mistaken for dogmatic assertion. Modern criticism establishes a norm as characteristic of Hellenis- tic comedy, and explains deviations from that norm in the Roman adaptations by certain known facts relating to the tradition of our Latin texts and the methods of composition followed by the Roman playwrights. The weakness of modern method may best be indicated by a consideration of the abnormal features — abnormal from the standpoint of modern critics — in several plays of Plautus; for this purpose I have chosen the Rudens, the Persa, and the Stichus. It is not my purpose to prove that the Rudens and the Stichus are not contaminated, or that the Persa is not from a Greek original of the Middle rather than of the New Greek comedy; I wish simply to illus- trate from these plays how certain features of dramatic technique are handled by modern critics without due regard to the demands of [Classical Philology XI, April, 1916] 125 126 Henbt W. Prescott the audience in the theater, to the limitations set by the scenic back- ground and the traditions of the literary type, to the restrictions which hamper the poet once he has sketched the main outline of his plot. In describing, therefore, the supposed abnormal features I shall briefly suggest the internal or external necessity which, in my opinion, made them inevitable in the Greek original; in this case they may not be explained as Roman defects due to contamination or retracta- tion; and they conflict with a theory of artistic regularity in Hellen- istic comedy which modern critics believe to have been induced through the influence of Euripides. The Rudens of Plautus is not conspicuously unlike other Roman comedies. It is, however, a diffuse play. Some of this diffuseness, as the licet- and censeo-scenes near the end of the play perhaps attest, may be Plautus' contribution.' But it is equally clear that no small amount of it is inherent in the Greek plot. A girl and her maid, carried off by a slave-dealer and his accomplice, are ship- wrecked; the discovery of this girl's status as a free citizen must be established by tokens which she has lost in the storm at sea; these tokens must be found at a relatively late stage in the action by a neutral or friendly person, or group of persons; of those shipwrecked with her none is friendly save the maidservant, and as this servant is needed for various dramatic and economic effects in the earlier scenes of the play, before the recognition can take place, the drama- tist may not use her to bring about the recovery of the tokens. It follows inevitably that an outside person, apart from the victims of the shipwreck, must accomplish the recovery; for this purpose a fisherman, Gripus, is invented, whose activity in the Roman play is limited to the latter half of the comedy. In the earlier action another slave, Sceparnio, served to connect minor chapters, to furnish some amusing effects; he now disappears; his function is completed; he was not available for the discovery of the tokens; Gripus, essen- tial to the recognition theme, becomes prominent in the last two acts of the comedy. Under these compelling circumstances a modern dramatist might make Gripus, if not thoroughly organic, at least less mechanically related to the action than the fisherman is in the Roman play. A playwright today, for example, might put in the mouth of Daemones, The Interpretation of Roman Comedy 127 the master of Gripus, in some of the opening scenes of the play, a casual remark to the effect that Gripus has gone out fishing, and that he wonders how the slave has fared during the storm; thus the audience would be duly prepared for the advent of the fisherman in later scenes. But in the extant play this casual remark, instead of being introduced early in the action is very mechanically brought in at vss. 897 ff. and immediately before Gripus' entrance; Daemones is very obviously lugged on the stage and haled off it (vss. 892-905) merely to provide this introduction of Gripus, with whom the previ- ous action has not made the audience acquainted. Now the removal of Sceparnio from the action, and the mechanical appending of Gripus for the purposes of the recognition theme, may well suggest that Plautus has combined parts of two distinct Greek plays; 1 and vss. 892-905, in which Daemones is so inartistically brought on and removed, may be Plautus' clumsy gluing of alien elements. But before any such supposition may become established fact or well-reasoned theory one must reckon with the situation that confronted the author of the Greek original; even the Greek author of the first three acts of the present play had to invent a character corresponding to Gripus. Once invented, it was difficult, in the nature of the plot, to make him an organic character; his connection with the main action, when he was first introduced, had to be loose; the mechanical introduction of Gripus in vss. 892-905 is indispensable to the needs of the audience, Greek as well as Roman, and, so far as it is mechanical, accords with frequent practice observable as early as Greek tragedy and Aristophanes. 2 Broadly stated, one is not 1 So Miss Coulter, Class. Phil, VIII, 57 ff., and cf . the references ibid., 57, n. 4. 2 The mechanical introduction of characters appears in the formulaic Kal nfyp 6/>t3 (commoner in tragedy than in Aristophanes) and ecce .... video with great frequency . Apart from the variations of this formula, note the obvious self-introduction of the parasite in Bacch. 573, and the patent address to the audience in Poen. 203^, where both the young women are known to the characters on the stage and haec est prior carefully distinguishes the one from the other for the benefit of the spectators. The phraseology of introduction has been considered by W. Koch, De personarum comicarum introductione, Breslau, 1914; for incidental comment cf. Fraenkel, De media et nova comoedia qu. sel., Gottingen, 1912, p. 59; Graeber, De poet. Attic, arte scaenica, Gottingen, 1911, p. 19; Flickinger, Class. Jour., X, 207 ff. A study of introductions, as a phase of the technique of comedy, will shortly appear, I hope, by Mr. D. M. Key, of the University of Chicago. It will be observed that the postponement of the intro- duction in the Rudens to vss. 897 ff. immediately before Gripus' appearance may be a concession to the needs of an audience that is not provided with playbills. 128 Henry W. Pbescott likely to appreciate properly these supposedly abnormal features of the Rudens until he has fully considered in all the Latin plays, in the remains of Greek comedy, even in Greek tragedy, the conditions under which inorganic or loosely attached characters are employed in the drama, and the means by which such characters are related to the action. A study of inorganic r61es would reveal differences in degree, broad resemblance in kind; possibly a difference of degree in Gripus' case might confirm a theory of contamination; but safe conclusions can be based only upon a comprehensive study of the entire phenomenon, not upon casual observation of Gripus' role in the Rudens. The Persa of Plautus is more obviously irregular than the Rudens. Operating with the same factors, largely, as do the students of con- tamination and retractation, but employing them to bring the date of the Greek original as near as possible to the time of Aristoph- anes and Euripides, Wilamowitz 1 has proved to the satisfaction of most modern students of comedy that the Greek model was a play of the Middle, not of the New, period. His argument from historical allusions is not relevant to my purpose; only his attitude toward supposed peculiarities of structure and character-treatment illus- trates the tendencies of modern method which I am examining. The play is primarily a slaves' play : a slave, plenipotentiary in his master's absence, intrigues against a slave-dealer; the slave- dealer owns the slave's sweetheart, a slave-girl; a second slave co- operates with the lover; a third slave, Paegnium, a puer delicatus, is loosely attached to the action to provide the comic byplay which relieves the general seriousness of the plot of intrigue. This general atmosphere of slaves temporarily liberated for the free exercise of their jovial and malicious propensities is very happily accentuated and preserved in the carousal which as an afterpiece follows the plot of intrigue elaborated in the first four acts; at this carousal the three slaves and the slave-girl sweetheart join in a triumphant convivial celebration in which the utter discomfiture of the slave-dealer reaches its culmination. But the demands of the intrigue require two free citizens — a parasite and his daughter; for the plot involves the palming off ' De tribus carminibus latinis (Index lect., Gottingen, 1893-94), 13 ff. The Intebpretation of Roman Comedy 129 upon the slave-dealer of a free woman as a slave. This pseudo-slave is to be ultimately claimed by her father, and the slave-dealer thereby put in jeopardy. The dramatist, in choosing a parasite and his daughter, has selected characters from the very lowest status of free citizenship, to that extent not entirely out of harmony with the servile status of the main characters, but as free citizens mildly disturbing the unity of atmosphere. This disturbing element is removed as soon as its necessary function in the intrigue is performed ; they are needed only for the intrigue, and their activity accordingly ceases when the trick is played. This mere statement of the poet's design, so far as realization may reveal the underlying purpose, should, in my opinion, meet sufficiently the objections of many modern critics. 1 They are disposed to insist that the parasite, who has been lured into active co-operation by the mention of appetizing foods and promises of perpetual feasts (vss. 140, 329 ff.), should be present at the concluding carousal. But clearly in the Persa, as in the carousal at the end of the Stichus, the presence of a free citizen would disturb the unity of a celebration designed to commemorate the emancipation, for the moment, of a group of slaves. The ancient audience was left to imagine that the parasite obtained his promised reward without dramatic realiza- tion of the feast that he had earned. 2 'Wilamowitz, op. tit., 20 ff.; Meyer, "De Plauti Persa," Comm. phil. lenenses, VIII, fasc. 1, 179 ff.; Miss Coulter, " Retractatio in the Ambrosian and Palatine Re- censions of Plautus," Bryn Mawr College Monographs, X, 39 S. 2 Economic factors are often disregarded in modern criticism: the addition of the parasite to the final carousal would, perhaps, increase the number of actors required for the production of the play. Wilamowitz' distribution of rdles (op. tit., 25 ff.) is not flawless: he has not provided for Sophoclidisca; and his division rests on the tacit presupposition that a vacant stage often marks an essential pause in Roman comedy (against which cf. Conrad, The Technique of Continuous Action in Roman Comedy, 1915) ; if the action is continuous at vss. 52 and 328, Sagaristio and Saturio may not be played by the same actor. The last two scenes show that at least five actors were required; the general structure indicates that the three heavy r61es of Toxilus, Dordalus, and Sagaristio each required a single actor; there remain five r61es, four of which are female or quasi-female (Paegnium) roles, that might be distributed among two or three actors; the same actor might play Paegnium and the mrgo; another actor might play Lemniselenis and Sophoclidisca (for, even if vs. 179 be assigned to the former, she need not have appeared on the stage) ; the only question is whether the parasite fell to a sixth actor (in which case he might have appeared in the carousal without increasing the number of the troupe) , or was added to the parts played by the actor who carried the r&les of Lemniselenis and Sophoclidisca. Against the second 130 Henby W. Prescott Wilamowitz is too familiar with the general weakness of comedy in respect to motivation 1 to lay much stress upon the defects in this regard of the amusing scenes in which the slave-boy, Paegnium, early in the play, is brought on the stage and elaborately presented to us in stationary lyrical scenes. The errands upon which Paegnium and the slave-woman, Sophoclidisca, are sent are quite futile, and serve simply as weak excuses for getting them upon the stage to amuse the audience and to lessen the seriousness of the more essential action. Yet the general indifference of the comic poets to motivation does not prevent Wilamowitz and others 2 from finding serious defects in another shorter passage of the play in which the weakness of motiva- tion is the most significant feature (as in the passage of the Rudens, vss. 892-905, discussed above). The intrigue in the Persa is com- pleted in two chapters: in one, the arch-intriguer purchases his sweetheart from the slave-dealer with money borrowed from his fellow-slave; in the other (incidentally, to repay the borrowed money) he tricks the slave-dealer into purchasing the parasite's daughter, a free woman, but represented to be a Persian captive. The dramatic effect is enhanced by carrying out both chapters in uninterrupted suc- cession, and the arch-intriguer remains on the stage dominating the situation (as, to a greater degree, in the intrigue of the Mostellaria) through both chapters (vss. 449-737). The parasite, who is needed only for the denouement of the second chapter, in which he must appear and claim his daughter as a free citizen, is introduced to us before the beginning of the entire intrigue (vss. 329 ff .) and withdraws to the house of the arch-intriguer (vs. 399), where he remains in hiding alternative stands the lack of harmony between the r61es of the parasite and the two women; in favor of it, stands the resultant economy and the structure of the play at vss. 305, 329, 752, 763; it may be that Sophoclidisca leaves at vs. 305 to assume the rdle of the parasite at vs. 329, and that the parasite leaves at vs. 752 to appear as Lemniselenis at vs. 763. 1 Op. cit., 22: "et intrant et exeunt personae plerumque soli poetae arbitrio obse- cutae, sin vero causam abeundi proferunt, ipso silentio peior est " and in his commentary on Euripides' Herakles, vs. 701: "es gehort zum Stile des griechischen Schauspiels, die Motivirung des Gleichgiiltigen zu verschmahen, und zum Wesen des antiken Publikums, Adiaphora als solche hinzunehmen und sich bei ihnen nicht aufzuhalten." Yet these interesting generalizations should be tested in careful studies of motivation in tragedy and comedy; cf . below, p. 144, n. 1. "Wilamowitz, op. cit., 21; Goetz, Acta soc. phil. Lips., VI, 300 ff.; Meyer, op. til., 172 ff.; Miss Coulter, Betractatio in ... . Plautus, 38. The Interpretation of Roman Comedy 131 during the whole of the first chapter and the greater part of the second. Thus for over 300 verses (vss. 399-726) he is entirely lost to view, and the scenic background, if he is within the house, pro- vides no means for his observing the earlier progress of the intrigue and knowing when the time for his own activity approaches. The audience, too, may well have lost track of him in such a long interval; certainly he himself must in some way be informed that the time for his arrival on the scene has come. With this dramatic problem before him the poet — and I must insist that the Greek poet confronted the same problem — devised the very mechanical action at vss. 711-30. The slave-dealer must be removed while the parasite is brought on and admonished; unable to remove the slave-dealer artistically, now that the threads of the action are pretty well spun out, the poet simply drags him off at vs. 723, leaving the parasite's daughter on the stage (naturally the slave-dealer would have taken her with him into the house) because she is needed in the subsequent action, and as mechanically dragging him back again at vs. 731 (so Daemones was dragged on and off at Rudens vss. 892-905). These much- discussed verses (vss. 711-30), therefore, are the dramatist's way of solving his practical difficulties; and those difficulties inhered in the Greek plot. Yet modern critics are so impressed by the obvious mechanism and general weakness of technique that they ascribe the supposed abnormalities to the botchwork of a later retractor, or insist that Plautus must have made over, at this point in the play, a Greek plot which, in its legal aspects, conflicted with Roman pro- cedure. 1 My own view is that the technique, however awkward, is explained so soon as we put ourselves in the place of the poet and the audience, and the Greek as well as Roman poet and audience. »A legal expert (Partsch, Hermes, 45, 613) ia not convinced by Wilamowitz' argument in this connection, and a layman finds it hard to believe that Dordalua is technically guilty under the circumstances. Is not the slave-dealer in jeopardy more because of a general prejudice against his class than because of any technical liability ? As my colleague, Professor Bonner, suggests, without being technically guilty Dordalus would be embarrassed by legal action, and that situation suffices for dramatic purposes and makes unnecessary the precise legal procedure which Wilamowitz posits as deter- mining the action in the Greek original. The fact that vss. 727-28 repeat vss. 467-68 is a textual problem ; in both places an accomplice is warned of his approaching activity in the intrigue, and both couplets may illustrate only Plautus' fondness for repeating himself (cf. Kellermann, "De Plauto sui imitatore," Comm. phil. Ienenses, VII, fasc. 1, 155, note on Persa 2). 132 Henry W. Prescott Rarely does modern criticism find in the portrayal of character any idiosyncrasy; but the peculiarities of the virgo in the Persa are used to support a view that the Greek original was specially exposed to the influence of the tragic poet, Euripides; the argument illus- trates the strength, in the minds of modern critics, of what I shall in later discussion call the presupposition underlying the modern interpretation of Hellenistic and Roman comedy. The characters of Roman comedy consist of stereotyped repre- sentatives of various trades and professions and slightly individualized domestic characters; they are in general realistic, but the realism of the portraiture is often modified under the stress of literary tradition or by immediate dramatic convenience. The choice of char- acters is determined by the inner necessities of the plot; in some instances external conditions seriously affect the selection. So, for example, the social conventions of Athenian life, combined with the rigidity of the scenic setting, which put the action of the plays in a public street and made interior scenes difficult, tend to eliminate from the comedies the respectable unmarried woman. The Persa is unique in its presentation of a virgo in an active rdle. Now this abnormal feature might well excite surprise, and lead any reader to recall the heroines of Euripidean tragedy, were it not that the unique character is immediately explained by the conditions of the plot. The plot of the Persa, as we have seen, requires a free and unmarried woman who shall be palmed off as a slave; no woman of the higher grades of Athenian society would lend herself to such a purpose; the dramatist chooses one from the dregs of society, and even she demurs to the task imposed upon her. Under these circumstances, if the portrayal of character in comedy is primarily realistic, we should certainly expect to find in the parasite's daughter a person totally unlike any other woman in the pages of comedy. What parasites' daughters were in contemporary society we have no means of knowing, but the general conditions of life and social custom as they affected women in the status of free citizenship would point to a limited horizon, a very narrow outlook upon life and its problems, especially before marriage. 1 1 Cf. Ferguson, Hellenistic Athens, 71 ff., for the difference between the cities of Greece proper and those of the outside Greek world in this respect. The Intebpbetation of Roman Comedy 133 In noting in some detail the talk and behavior of the virgo we should remember that in the second of the two scenes in which she appears she has been taught a part (vss. 379-81). Such inconsist- ency as may be apparent between her seriousness in the first scene and her cleverness and apt repartee in the second is thereby ex- plained: "necessitate me mala ut nam facis" (vs. 382); even apart from this reasonable explanation of the slight contradiction, any dramatist in Hellenistic comedy is prone to abandon consistency of character if the immediate needs of the action are consequently advanced. 1 In the first scene (vss. 329 ff.), in language and manner strange to the reader of New comedy, who has become familiar largely with mercenary courtesans or irate wives, the virgo protests against the use which her father purposes to make of her. He is selling her to fill his belly (vss. 336-38); though poor, they should better preserve their good name than become rich at the expense of their reputation (vss. 344 ff.); Mrs. Grundy is a dangerous enemy (vss. 351 ff.). Her father retorts that his own appetite is the first consideration, and that the sale is not a real one; but she does not like even the pretense (vss. 357 ff.). The parasite resents wisdom in a daughter, and regards it as a weakness of her sex; the only weakness, she replies, is in letting evil action go unnoticed (vss. 365-70). She yields only to his authority as father, and points out that if the family gets a bad name her marriage will be difficult to arrange. The parasite, however, with a thrust at the evil times, declares that a dowry, not a good name, counts in marriage; and, reminded by her of his poverty, he finds in his stock of funny stories incal- culable wealth sufficient to achieve her marriage even with a beggar (vss. 383 ff.). In the second scene (vss. 549 ff.) she carries out her part in the trick with amazing cleverness. She comes on with a slave who poses as her attendant envoy from Persia; the two pseudo-foreigners are engaged in conversation as they enter. The attendant inquires whether she is not impressed by the splendor of Athens; she replies that she has seen only the external beauty of the town; the character of its citizens is still unknown to her, and she withholds judgment; 1 Cf. Legrand, Daos, 309. 134 Henry W. Prescott like a street-preacher delivering a diatribe she moralizes, lists ten deadly sins, 1 and declares a single wall to be a sufficient state of preparedness if the citizens are innocent of these sins, but a hundred- fold wall to be insufficient if a town is corrupted by such vices. The slave-dealer, considering the possibility of purchasing her, quizzes the virgo; her answers are ingenious; they satisfy the purchaser without committing her or her accomplices; she misrepresents her actual condition as daughter of the parasite in only one respect: she refers to herself as a slave, but otherwise every response fits her real situation as the parasite's daughter, with no little wit at the expense of the parasite for the audience to enjoy. She expects her parents to redeem her but does not object to a brief period of slavery (vss. 615 ff .) ; yet she weeps over her temporary plight. Dordalus inquires her name; it is Lucris, an auspicious name from his stand- point. Where was she born ? Her mother told her "in a corner of the kitchen." Reminded that he means in what country was she born, she insists that she is without a country except that country where she happens to be; the past is gone (vss. 630-38) ; and pressed for an answer she contends that Athens must now be her country (vs. 641). Was her father a captive? No, not a captive, but he lost what he had. What is his name ? His name is Miser and hers Misera. 2 What is his social standing? Everybody likes him, slaves and freemen, and she warns the slave-dealer that her father will ransom her, his friends will stand by him, even if he has lost his property. The slave-dealer is completely won over, of course too easily; the comic intrigue regularly represents the object of the intrigue as a gullible fool. The entertainment and the dramatic effects afforded by the virgo are admirable, of their kind; in the first scene her modesty and idealism are in amusing contrast with the coarse practical wis- dom of her father; and in the second scene she is contrasted implicitly 1 If both passages come from the Greek originals of the two plays, anybody who is seeking the date of the Greek model of the Persa might well note that the ten sodales of Persa 561 are matched by the sodales of Merc. 845 (six in 845, ten in 848-49). 1 Wilamowitz (op. cit., 25) finds a Euripidean background in the answer of Orestes, when he is asked his name, in Iph. 500: rb /liv dlxawv Svarvx^s KaXol/tcB' S.v. Cer- tainly the diction may be an echo of Euripidean phraseology, but the general idea in "nunc et ilium Miseram et me Miseram aequom est nominarier " is on the same plane with Gelasimus' "Famem ego fuisse suspicor matrem mihi" (Stick. 155). The Interpretation of Roman Comedy 135 with the slave-dealer, at whom her moralizing deals some sharp thrusts. The serious-mindedness which is her permanent char- acteristic is sustained, with advantage to the intrigue, in the second scene; her quick wit is perhaps foisted upon her to some extent by the dramatist for his immediate needs. The serious-mindedness, the moralizing, are the qualities of her class, affected by the secluded life of unmarried women in contemporary society; doubtless they are exaggerated for dramatic purposes, as are the qualities of the cook, the soldier, the courtesan, in comedy; but why need we turn to the tragic heroines of Euripides, rather than to real life, for an explanation of the virgof Must she be a tragic heroine simply because she is serious-minded ? And where in Euripides is such sustained prudish- ness to be found in the tragic heroine ? The individual sentiments, to be sure, in content and phrasing may be like the sententiousness of Euripidean characters, hut the influence of the tragic poet upon diction and style is pervasive throughout the Hellenistic period; the style of individual sententiae is hardly different from those that occur in the Mercator and Trinummus. 1 In brief, the inner necessity of the plot makes the virgo indispen- sable. Her essential features are those of the virgo, probably, in contemporary life, with some exaggeration; and though the char- acter is required by the Greek plot, we must grant in this case the possibility that Plautus expanded suggestions in his Greek original with a view to portraying a staid Roman virgin from his own environment. 2 1 Cf. Leo, Plant. Forsch?, 136. ' It is hardly fair to Wilamowitz to separate his confessedly weak arguments based on the technique of the Persa from the pivotal point of his discussion, but I wish simply to use the material to illustrate the general attitude of modern critics toward various phases of dramatic form in Roman comedy. On his main point, that vs. 506 presupposes an independent kingdom of Persia and dates the Greek original before 338 B.C., I may say that not merely the fictitious nature of the situation but the purely fabulous "Goldtown" which the Persians capture suggest to me a Utopian Eldorado rather than any accurate reference to contemporary history. I sympathize with the views of Meyer, op. tit., 183 ff. (but not with his argument, ibid., 186 ff.). The logical relation of Leo's note on Diogenes (Hermes, 41, 441 ff.) to the argu- ment of Wilamowitz should be thoroughly understood. Leo says that if Wilamowitz has proved his case, vs. 123 is a specific reference to Diogenes. On the other hand, it should be clear that if Wilamowitz has not proved his case, Leo's evidence, which is largely from Leonidas of Tarentum, probably a younger contemporary of Menander, would only establish the fact that at the time of the New comedy the equipment ascribed to the Cynic in vss. 123 ff. was attributed to Diogenes and other members of the school. 136 Henry W. Pbbscott Leo's masterly analysis of the Stichus 1 has doubtless convinced many students of comedy. Certainly the extraordinary structure of the piece may be satisfactorily explained as a combination of parts of three Greek plays; these three parts might be called "The Faithful Penelopes," "The Discomfiture of the Parasite," "The Slaves' Carousal." The Roman play mechanically joins these alien elements by attaching to the first part a parasite, whose hopes of a dinner are twice frustrated in the second chapter, and to the second part a slave, Stichus, who arranges the revel with which, as in the Persa, the play concludes. That the parasite and the slave are inorganic roles, that the play as a whole completely disregards unity of persons, are incontestable facts. The only question is whether such structure is inevitably Roman and Plautine. The modern critic denies that a Greek author, and particularly Menander, to whom a didascalic notice attributes the Greek original, was capable of this artless mechanism. 2 There is only one fact in Leo's analysis which I should qualify. Leo maintains not only that the first appearance of the parasite, Gelasimus, is not motivated (vss. 155 ff.), but that his appearance before the houses of the brothers is in flat contradiction of an essen- tial presupposition of the plot. Of course my own main contention is that it is idle to deal with the matter of motivation until this aspect of dramatic technique in its entirety has been properly studied, not only in comedy, but in tragedy. Accepting, however, for the 1 Cf. Nachr. d. dotting. Gesellschaft (1902), 375 ff. 2 "Menander und Plautus sohreiben beide fur die Biihne, aber Menander aus einer grossen Kunstentwicklung heraus, Plautus fur die kilrzlich entwickelte Buhnen- bedttrf ness eines kunstf remden Publikums ; ihm darf man die Umgestaltung Menanders auf grSbere Biihnenwirkung hin nach der Lage der Dinge wohlzutrauen" (op. tit., 377). "Was die Form angeht, so liegt Menanders strenge und konsequente Kunst vor Augen; sie vor allem gibt den Massstab ftir die Treue der rdmischen Bearbeitungen " (Gesch. d. r/)m. Lit., I, 108). Such statements illustrate the broad vision of the critic, and if limited to Menander may be sound generalization. But the question arises whether the absolute uniformity of art here attributed to Menander may not have been violated, in at least one play, by an author who left over 100 comedies; modern critics seem to forget the enormous productivity of the comic poets in the Hellenistic period; how complete regularity may we expect in authors like Antiphanes, credited with 260 plays, Alexis with 245, Philemon with 97? Furthermore, in his critical studies of comedy Leo seems to regard all the Greek writers of comedy in the New period as issuing from a great "Kunstentwicklung," for clearly whatever is not "Kunst" in Plautus is, in Leo's theory, Roman. The Interpretation of Roman Comedy 137 moment Leo's standpoint, I should say that the implicit motivation of the parasite's appearance is strong. The presupposition in question was not known to the audience at vs. 155, when Gelasimus first appears, unless it had been clearly stated in a prologue now lost to us. A modern reader has to recon- struct from vss. 214, 267, 372, 462, 584, 628, a very important fact: from these verses it is clear that the two brothers had been forced into their triennium of foreign travel by previous riotous living under the efficient guidance of the parasite; their commercial activity is an attempt to recoup their lost fortunes; the relations, therefore, between the parasite and the households of the two brothers are far from friendly during the action of the play. During the triennium the parasite has never been entertained at the homes of the brothers, and he is surprised that he, of all persons, should be selected as the messenger of one of the sisters and sent to the harbor for news of the absent husbands (vss. 266-68). Under these circum- stances it is even more surprising that without any explicit reason he should voluntarily appear before the houses of the brothers at vs. 155; he arrives before the servant, Crocotium, can fetch him; and the futility, from a dramatic standpoint, of his function as mes- senger is immediately clear when in the next scene the news is actually brought by Paegnium, permanently stationed by the sisters as a lookout at the harbor. It is obvious that the author, Greek or Roman, wanted Gelasimus on the stage at this place and time, and lugged him on without any artistic manipulation of circumstances. Yet, if we are to regard motivation as essential in the dramatic art of the Greeks and Romans, I should suggest that the monologue of Gelasimus at vss. 155 ff. sufficiently explains his presence: he is reduced to starvation, and forced to sell all his scanty property; under these conditions where should he more naturally appear, as a last resort, than at the houses of the two brothers, where, in his palmy days, he had been a welcome visitor ? But this fact is merely implied, and I have no reason to question Leo's general conclusion that the handling of the situation is abso- lutely mechanical; both Gelasimus and Stichus are loosely attached to the action; they are in different degrees inorganic characters; and it may be remarked, in passing, that in only a slightly less degree 138 Henry W. Prescott the parasite, Ergasilus, in the Captivi, from a comparison with whom Leo draws such important conclusions, is an inorganic character. 1 The reason for the existence of these two inorganic r61es (and the weakness, consequently, of Leo's entire theory that three Greek plays furnished the three chapters of the Stichus) is immediately apparent just so soon as we put ourselves in the place of a playwright who had devised the broad exposition of the first two scenes (vss. 1- 154) as a complete introduction of a Greek play. For Leo admits that this exposition, with its contrast between the two loyal wives, and the further contrast between the two Penelopes and their father, is thoroughly Menandrian; nor is it demonstrably incomplete; he admits, too, that of the three Greek plays which he posits as fur- nishing the three chapters of the Stichus he can construct the plots of the last two but is absolutely unable to think out to the end the Greek play from which came the Menandrian exposition of the Stichus. Can there be any other conclusion, unless Menander's imagination was more fertile than Leo's, than that Menander himself, having constructed the exposition which we find in the Stichus, would be forced to develop inorganic characters and a relatively inorganic play as the result of such an exposition ? Given, as the exposition, two loyal wives separated from their husbands for a triennium, and urged by a practical father to assume that their husbands are lost to them and to marry again without further delay, what dramatic action can develop when these two husbands return and become reconciled to their wives and father- in-law ? Leo is much disturbed that no visible reconciliation takes place, that the elaborateness of the exposition is not justified by any subsequent action; the sisters do not appear on the stage again after the exposition; the reunion of husbands and wives takes place off-stage; the reconciliation with the father-in-law is stated in the case of one brother, more visibly presented in the case 1 For, however well explained his presence in the opening scene before Hegio's house, the relation of Ergasilus to the main action of the Captivi is so loose that Ladewig and Herzog long since suggested contamination or Plautine invention to account for the inorganic r61e; these suggestions have been long discarded; yet the difference between Ergasilus and Gelasimus is one of degree rather than of kind. The degree may- be important, but its importance can be estimated only after a complete study of inorganic rfiles in comedy, not after a casual comparison of the two parasites in the Captivi and the Stichus. The Intebpretation of Roman Comedy 139 of the second. In brief, though all the facts of Leo's analysis might harmonize with the conclusion that the Stichus is a composite of three Greek plays, it is equally clear that the Menandrian exposi- tion, if complete, would compel the Greek author to attach a char- acter like the parasite to the first chapter of the action in order to provide the action of the second chapter, and if so trivial a character as Gelasimus were invented, nothing remained but the invention of a second inorganic character like Stichus 1 to give the play the requisite length, however much the unity of persons is thereby disrupted. There is nothing inevitable in Leo's conclusions; and if papyri from Egypt should ever confirm his conclusions, would it not be a tribute to his intuition rather than to the soundness of his argument ? Again, however, we are not interested primarily in the theory of contamination. Here, as in other plays, the critic has isolated as peculiar features of a single play certain supposed weaknesses; they are defects in motivation of entrance, in the organic relation of char- acters to action, and of exposition to subsequent dramatic develop- ment. Leo, to be sure, compares and contrasts the parasites of the Stichus and of the Captivi; but he had no complete study of any one of these three aspects of dramatic technique upon which to base sound conclusions. May I suggest, if only again by a single parallel instance, the need of less casual procedure in the handling of the technique of comedy? A very important item in Leo's argument is the wastefulness, in the present text of the Roman play, of the admirable exposition, both of the general situation and of the characters, in the opening scenes of the Stichus. Quite apart from my suggestion that this apparently useless introduction and the inorganic action that follows are explained by the conditions of the exposition itself, there is clear evidence that the lack of close organic connection between exposi- tion and main action is not peculiar to the Stichus, and is demon- strably Greek rather than Roman in its broad aspects. 1 The mechanical inlay which Leo finds at vss. 419-53 and attributes to Plautus' efforts to attach Stichus to the action of the play might have been better in the original. If Stichus was really part of Menander's Adelphoe, the passage vss. 435-53 (vss. 446 ff . reveal Plautus' hand) may be a substitute for a lyrical intermezzo by the music girls (hasce 418) in the Greek play; or so at least I should expect the many searchers after survivals of the Menandrian chorus in Roman comedy to suggest. 140 Henry W. Prescott The Mostellaria of Plautus, from the Greek of Philemon, is admitted by Leo 1 to be thoroughly Greek in all essential features of the action; nor does he contend that any of the exposition is Plautine save the solo-song of Philolaches, which, he thinks, in the Greek original appeared as a monologue; of contamination the play is as innocent as any Roman comedy can be. The four introductory scenes of the play constitute the most elaborate exposition in extant comedy: in the first scene two slaves in dialogue reveal the general situation — the riotous life of a son under the direction of a slave in the father's absence; in the second, the son himself in song reveals his weakness of character; in the third, the young man's slave-girl sweetheart, now liberated with borrowed money, is sharply con- trasted with an old beldam, her servant, and the effect is admirably portrayed in the changing moods of the eavesdropping lover; a fourth scene provides the general atmosphere of the whole situation — a boon companion and his sweetheart join the other pair of lovers in a brief, broadly humorous, lyrical intermezzo. Yet after this extended exposition of character as well as of situation the persons intro- duced to us in these scenes practically disappear for the rest of the action; Tranio, one of the slaves in the first scene, does become the arch-intriguer and dominates the later action, but the lover and his sweetheart, whose characters have been so fully delineated, are removed from the stage, and the subordinate boon companion, briefly presented in the fourth scene and at the beginning of the main action, is merely lugged on at the end of the play as a homo ex machina to cut the knot. The removal of the hero and the heroine, as we suppose them to be from the exposition, is cleverly devised; in the Stichus the wives are removed from the action only by the dramatic necessity, by the impracticability of developing dramatic action through their presence. But I should be disposed to assert that these expositions of two different plays supported a view that two Greek poets, Menander and Philemon, were so interested in char- acter per se that they disregarded close interrelation of exposition and main action to indulge in the portrayal of persons essential to the situation but irrelevant to the subsequent action. 1 Leo, Gesch. d. r&m. Lit., I, 110 ff. The Interpretation of Roman Comedy 141 As my choice of illustrations indicates, I am concerned only with those general aspects of the form of Roman comedy from which modern critics obtain their notions of the relation of Plautus and Terence to their Greek originals; these critics are intent upon the obvious and laudable task of reconstructing the history of ancient comedy, and of placing, not only Plautus and Terence, but the Greek authors whom these Roman playwrights translated, in their proper historical relation to one another and to antecedent comedy. The problem that confronts these modern scholars, who, in view of the fragmentary remains of Hellenistic comedy, may well seem ambi- tious in their aims, is a difficult one. The plays of Aristophanes present an incoherent satirical burlesque, provided with a chorus as an organic part of at least the first part of the play, and lampoon- ing public men, public policy, and the general trend of ideas and cus- toms in contemporary Athens — -a local product, instinct with the life of the Greek metropolis of the fifth century. The comedy of the Hellenistic period exists only in fragmentary form, through which the content is only vaguely discernible, the structure even more difficult to determine. The 26 Roman plays adapted from this Hellenistic comedy present a coherent drama of private life, of sentiment, and of intrigue, without a chorus, in which a generalized picture of contemporary society has replaced the fantastic treatment of local problems. How are these two diverse types of comedy re- lated to each other, if related at all ? Ancient theory, expressed in a number of Byzantine documents, and perpetuating with later accretions academic opinion that may in some respects be as old as the school of Aristotle, 1 solved this ques- tion by the assertion that financial pressure led to the elimination of the chorus in Old comedy, that political conditions made impracti- cable the open criticism of men and events, the implication being that these two causes suffice to explain the development of an incoherent satirical burlesque into a well-organized realistic comedy of manners. Making all allowance for the facts, that in some plays of Aris- tophanes the chorus falls into the background or almost completely disappears, that relative unity is occasionally discoverable even in 'Kaibel, "Die Prolegomena IIEPI KOMftlAIAS," Abhandl. getting- Gesell., II/4 (1898). 142 Henry W. Pbescott plays in which the chorus is prominent, that in his latest plays Aristophanes made use of sentimental legend in which the exposure of a child and its recognition were essential features, modern criti- cism, with some reason, refuses to admit that ancient theory satis- factorily accounts for the sharp contrast in both form and content between Aristophanic plays and the extant comedies of Plautus and Terence. It asserts as an incontrovertible fact 1 that at the turn of the fifth century the younger Greek tragedy, represented chiefly by Euripides, exerted a potent influence upon the form and content of comedy, which, once the chorus was removed, took over the coher- ent structure of Euripidean tragedy and perpetuated without impedi- ment the ideals that the tragic poet, hampered by the conventions of his literary type, could only faintly realize. Euripides, yearning to portray realistically contemporary life, must content himself with making of the demigod Orestes a very ordinary human being; Hellenistic comedy, free from the restrictions of tragedy, and relieved of the chorus as an organic element, easily conformed to a demand for realistic portrayal of private life and attained artistic unity. Roman comedy, therefore, reflecting the Euripidean form of Hellen- istic models, reveals a serious framework of well-knit action, with comical appurtenances, and a happy issue; its artistic unity and much of its content are an inheritance from Euripidean art. The features of Roman comedy which establish this fundamental presupposition are briefly these: 2 The plays are serious, the comic elements often detachable; remove the parasite from the Captivi, change the issue, a tragedy results. The emotions exhibited and excited by Roman comedy are, mainly, those proper to tragedy rather than comedy. The plot of intrigue is anticipated in several plays of Euripides. The exposure of children at birth and their later recognition is a tragic theme and situation. The intimate life and the domestic characters have little or no background in Aristophanes but are suggested in the tendencies of Euripides which Aristophanes delighted to ridicule. The plays of Euripides, in which the chorus is 1 Leo, Gesch. d. r8m. Lit., I, 100 ff. • Ibid., I, 104 ff. For a comprehensive statement of the case from the standpoint of the historian of Greek literature cf. Christ-Schmid, Gesch. d. griech. Lit., II/l*, 26 ff. References on the details are deferred to the fuller discussion of the theory in a subsequent paper. The Interpretation of Roman Comedy 143 often somewhat detachable, fall into six or seven coherent chapters, separated by choral songs; modern critics enjoy finding in the vacant stages of Roman comedy the deserted abodes of a tragic chorus, these empty spaces setting off well-defined chapters of action and distinguishing a general unity of form. 1 The most conspicuous feature of the technique of Roman comedy is solo-speech and solo- song; Euripides not only extensively cultivated solo-song by the actors, but more conspicuously than the other tragedians resents the impediment of the chorus, and on occasion puts in the mouth of a character a quasi-soliloquy in spite of the presence of the chorus, or rarely removes the chorus and resorts to monologue. Of these solo- speeches in comedy the prologues of many plays of Plautus, both in the style in which they set forth the plot and in respect to the char- acters in whose mouths they are put, are Euripidean. Aside from these essential elements of form and content Roman comedy in a variety of lesser features, in the use of various dramatic devices, in general sententiousness, in digressory moralizing upon social conditions and the proposal of social reforms, in attacks upon social groups, is supposed to reflect the tedhnique and the substance of later tragedy. Having thus confirmed the basis of its procedure by a substantial amount and quality of evidence modern criticism finds in the Roman plays, more conspicuously in the comedies of Plautus than in those of Terence, 2 striking exceptions to the uniformity of structure and content demanded by its Euripidean theory. The Casina, for example, is a broad farce. 3 Various plays operate with inorganic 1 Difficulties in the assumption that vacant stages in Roman comedy with any regularity mark essential pauses are indicated by Conrad, op. cit., passim. 2 The differences between the six plays of Terence and the twenty plays of Plautus have probably contributed largely to the development of the methods of modern criticism. Terence reveals a relative regularity in structure; Plautus has many vagaries. The younger poet is supposed to be more closely adapting his Greek originals; Plautus' vagaries are supposed to be Roman. But Terence is two-thirds Menandrian; may not his relative uniformity be merely the relative regularity of a single Greek author, and Plautus' variety, though sometimes Roman, in general the individual variations of a number of Greek playwrights ? Modern interpretation of comedy is hampered, from my standpoint, by this concentration upon the uniformity of Terence-Menander; the diversity of Plautus may be the diversity of Hellenistic comedy in the large. » Mainly because of its farcical character and the preponderance of song the Casina, in modern theory, becomes largely a Plautine composition (Leo, Gesck. d. rOm. Lit., I, 126 ff.). It is not clear how Leo would reconcile his two opinions that the songs 144 Henry W. Pbescott characters. The chapters of action are not always artistically joined. Entrance and exit are often weakly motivated, or not explained at all. 1 There are logical contradictions in the facts of the plot. Obviously, if the main presupposition is sound these irregularities and unevennesses must be explained in accord with the presupposition of general dependence upon Euripides. The explanation is found by the critics in two known conditions, one in the Roman methods of composition, the other in the transmission of the Roman texts. An accumulation of weaknesses in one play, especially a combination of two plots of intrigue directed to the same end and accompanied by contradictions of fact and related weak- nesses of technique, is explained as due to the adaptation of two or more Greek plays in one Roman copy, to contamination. Isolated defects in various plays are justified as the result of corruption in text-tradition; these were promoted especially by the reproduction of Roman comedies in the generations after Plautus and Terence, such reproductions leading to the revision of the original text by the hands of stage-managers. of the Casino, are evidence of Plautine workmanship, and that the lyrical parts of the Persa, if Wilamowitz' view of the play is right, require us to consider "ob nicht auch im Original [i.e., of the Persa] mehr als bei Menander und Philemon gesungen wurde " (op. cit., I, 120). Because of its complexity I have not attempted, in this brief accou nt of modern interpretation, to state the relation of the problem of the cantica in Plautus to the historical development of comedy. 1 The broader aspects of motivation in tragedy and comedy must be considered before one may estimate the significance of apparent resemblances between Euripides and the New comedy (cf. C. Harms, De introitu personarum in Euripidis et novae comoediae fabulis, Gottingen, 1914). Dramatists may independently develop similar methods of motivation. A stereotyping of dramatic devices may have arisen naturally at the dramatic festivals at Athens which would lead to resemblances in such features between tragedy and comedy without necessarily indicating predominant influence of one type upon the other. The rigidity of scenic background and the domestic setting which New comedy has in common with Euripidean tragedy might produce some com- mon devices of motivation which would not, therefore, establish a theory of Euripidean influence. And finally, Euripides as the precursor of the Hellenistic period would naturally anticipate New comedy in many respects without any direct influence upon the type; so, for example, in the matter of motivating entrance by the emotion of fear (Harms, op. cit., 28 ff.), note the general features common to all the tragic and comic poets, and the single trait which Euripides and New comedy have in common, the emotional exordium. Is this exclusively due to the influence of Euripides, or is it the emotional elaboration of the later epoch which Euripides anticipated ? And if there is resemblance in the diction and style of this emotional exordium, is this kind of influence pertinent to a theory that Euripides at an early date determined the form and some of the content of New comedy ? The Intebpbetation of Roman Comedy 145 The result of this modern theory and method is a certain neatness and dispatch in the interpretation of Roman comedy. Terence is admitted to be an artistic contaminator; the commentary of Donatus supplies the evidence. But in Plautus everything artis- tically satisfying is Greek in origin, everything defective and weak is Roman botching, whether of Plautus himself or a later Roman hand. This differentiation of the two dramatists is not at odds with what we know of the different nature and entourage of the two poets. Modern theory has progressed so far that now about one-half of the Plautine corpus is supposed to show the injurious effects of contamination; the scope of retractation is undefinable, affecting various plays in different degrees. Either retractation or contamination is safely used to explain conditions in our texts when the critic rests his case on a substantial basis of evidence. For example, retractation is often solidly estab- lished when the text presents duplicate passages; but the critic enters upon more dangerous ground when mere weakness of technique starts the application of the principle. Contamination is a rela- tively sound explanation if the critic finds in a given play essential contradictions of fact combined with a double intrigue, each part of which is directed to the same end, and both parts of which are mechanically affixed to each other, as may be the case in the Poenulus and Miles. But I venture to protest strongly against the applica- tion of either principle, and against the presupposition of modern criticism, when they operate exclusively with supposed weaknesses of dramatic technique, or with any apparently abnormal features, that find immediate explanation in the conditions of the dramatic plot, in the needs and demands of the ancient audience, Greek as well as Roman, and in the peculiar arrangements of the ancient stage and theater or in the conventions established by known literary tradition. The casual and incomplete treatment of large problems of dra- matic technique in Roman comedy is due in some measure to the concentration of modern students upon the important questions raised by contamination and retractation; the narrowness of vision induced by such concentration is further increased by intensive studies of single plays which seem, to individual critics, to reveal the 146 Henry W. Pkescott effects of these two factors in the composition and the transmission of the Latin texts. But even more effective than the consequences of this concentra- tion upon a limited amount of text, and upon two distinct problems raised by the text, has been the fundamental presupposition which dominates the higher criticism of Hellenistic comedy and the Roman adaptations. The assumption, or, as modern criticism holds, the incontrovertible fact, that Euripidean tragedy exerted a potent influence upon comedy at the turn of the fifth century at once establishes a rigid norm and closes the minds of the critics to the possibility that many of the defects of form in Roman comedy are Greek in origin and natural survivals of the incoherence of earlier stages of the Greek type. Modern theory, in spite of its emphasis upon Euripidean influence, cheerfully grants that the Hellenistic type in many respects continues and develops important features of Aristophanic comedy; 1 the critics maintain simply that later comedy inherits more of the essential characteristics of one parent than of the other. Under these circumstances there are clearly other possible explanations: if the evidence warrants it, Hellenistic comedy may derive neither from Aristophanic comedy nor from Euripidean tragedy nor from the marriage of both, but from a different source which combined a relative unity of structure with characters and incidents inherent in comedy rather than tragedy; or it may be that later Greek comedy, like so many other Hellenistic types, was subject to a variety of influences, among which Euripidean tragedy is less significant than modern critics suppose. Certainly many aspects of Hellenistic comedy which resemble corresponding features of the later tragedy are sufficiently explained without resort to the conten- tion that such influence was exerted overwhelmingly at the turn of the fifth century. Through mythological travesty, as the critics admit, tragedy influenced comedy at a much earlier period; later, the pervasive influence of Euripides upon Hellenistic poetry and the direct influence of the tragedian upon individual poets like Menander inevitably affected various Greek playwrights in different degrees. But this statement of the case is far different from a view that 1 Leo, Gesch. d. rom. Lit., I, 104 ff.; and (correcting the misapprehension of Suss) Plaui. Forsch.', 113, n. 2. The Interpretation of Roman Comedy 147 Euripidean art was so dominant an influence early in the develop- ment of Hellenistic comedy that a regularity of artistic form was established, deviations from which may be explained only as due to Roman corruption. If this presupposition without being entirely demolished is appreciably weakened, the foundations of modern interpretation are unsettled, the criteria of contamination and retractation are subject to revision, and the problems of dramatic technique must be regarded from a different standpoint. In the next paper, therefore, I shall endeavor to review without prejudice the evidence bearing upon the antecedents of Hellenistic comedy. Obviously the interpretation of dramatic technique in Roman comedy must depend upon the conclusions of such a study if the evidence warrants positive conclusions. University of Chicago