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. THE ANTECEDENTS OF HELLENISTIC COMEDY By Henry W. Pbescott VI Even if, as I have suggested, 1 the indebtedness of comedy to Euripides in the material of plot is not so large in amount or so significant in kind as modern critics have represented, it still remains quite possible that in form comedy is dependent upon Euripidean tragedy. Into a mold provided by a different type of literature comedy may have poured a new content. Indeed it is undoubtedly the striking contrast between the looser epirrhematic and episodic structure of Aristophanic comedy and the organic coherence of Hellenistic comedy, as seen in the Roman copies, that has led modern scholars to reject the ancient theory in the prolegomena and to stress heavily the broad resemblance, in point of unity, between later comedy and Euripides. Nor do the variations in the structure of Aristophanic comedy effected by the postponement of the agon to the second half of the play, and by the diminished r61e of the chorus in the Ecclesiazusae and the Plutus, very appreciably lessen the gap in this respect between Aristophanic and Hellenistic comedy. Taking organic structure in the broadest sense, before we imme- diately accept the Euripidean theory, must we not ask ourselves, in view of the fact that there is no full and specific relation between the plots of Euripides and those of the later comedy of manners, whether the comic plot of the later period, without any immediate inter- vention of earlier or contemporary tragedy, does not bring into the comic drama at once a degree of coherent structure that the mere themes of Aristophanic comedy made impossible in the scurrilous plays of the fifth century ? These comic plots of the fourth and later centuries are not homogeneous; the twenty-six Roman plays reveal a variety of plots, and the Greek titles and fragments increase this variety. The comedy of manners, with which alone we are at present concerned, may have been a renascence of one kind of Sicilian-Attic comedy, or it may have issued immediately from the private life iCP, XIII (1918), 113 ff. [Classical Philology XIV, April, 1919] 108 The Antecedents of Hellenistic Comedy 109 of the fourth century. Its precise origin does not matter for our present purpose. Of its various plots a common one, which we may use for illustration, is the story of a young lover prevented from indulging his love for a courtezan by obstacles, usually of a pecuniary sort; the lover himself, or a slave, or parasite obtains the required financial help, usually through some swindling intrigue, and, often further assisted by the discovery of the courtezan's free birth, attains his end. 1 That such a plot is the issue of any slow literary evolution is difficult for me to believe. The broad outline of this story offers in itself a beginning, middle, and end, with obstacles and means of solution that are easily varied and multiplied. It is quite superfluous for tragedy to superimpose upon this type of plot a general coherence and logical organization which it already possesses. It is, on the other hand, quite true that mythological comedy, 2 which had prevailed in the period immediately preceding the vogue of the comedy of manners, had in many instances acquired an organic structure by being a travesty of well-organized tragedy; and one cannot easily say how conscious of the advantages of an organic form comic poets may have become through constant witnessing of tragic dramas as well as by intermittent perversions of tragic plots. My point is merely that the material of the comic plots is almost entirely independent of tragedy, and that the unity, in a broad sense, is possibly furnished, without any long period of artistic development, by the simple realistic tale of human experiences. 1 The theme is a variant of the eternal commonplace which Post (Han. Stud. Class. Phil., XXIV [1913], 112) reduces to a formula. The broad similarity to the plots of later Greek romances is obvious. The romances themselves, however, are often called dramata by their authors; this implication of dramatic influence upon the romances makes it difficult to assert that early prose fiction, no longer extant, contributed to the material of comedy. But the possibility is always open; for interesting reflections cf. Mahaffy, Greek Life and Thought, 118 ff.; Bousset, Ztschr. Mr die neutestamentl. Wiss., V (1904), 18 ff.; Mendell, CP, XII (1917), 161 ff.; and specially Thiele, Hermes, XLVIII (1913), 536, n. 1, 539, n. 1. 2 The salient facts regarding mythological comedy seem to me to be that (1) oral tale and epic must have brought some unity into mythological comedy before tragedy exerted any influence; (2) that the influence of tragedy was exerted probably as early as the time of Epicharmus; (3) that mythological comedy was probably infinitely varied, (a) as illustrated by the Plutus, which suggests the loose unity provided by the application of a legend with allegorical implications to contemporary life, (6) by the Dionusalexandros, in which fantastic perversion of myth could hardly have promoted unity at all, (c) by the Amphitruo, which shows the high degree of unity attainable through the fusion of a tragic plot with a comedy of errors. 110 Henry W. Pkescott Politics, literature, and philosophy did not supply Aristophanes with themes that were inherently dramatic and easily organized into effective dramatic chapters, but typical experiences of real life, such as the recurrent plots of New comedy reveal, hardly need the impress of tragedy before they can assume at least a considerable degree of organic unity. However abstract and a priori this reasoning is (as it must be in the dearth of positive evidence), it is interesting to observe that before the middle of the fourth century the general coherence of the comedy of manners is recognized by a comic poet, Antiphanes; the invention of the presuppositions, of the facts of the plot, of the exposi- tion, and of the catastrophe, in comedy as well as in tragedy, he seems to view in a detached and conscious fashion and to describe them in terms that to some extent suggest an almost academic attitude toward dramatic structure and an apparatus of technical labels. He is referring to the advantages of tragedy in dealing with stories familiar to the audience, supplied with characters whose names and expe- riences are already known, and in having the mechane available for emergency; in contrast therewith he puts the comic poets who have to invent everything — new names, presuppositions, plot, catastrophe, exposition. It should be clearly understood that the fragment refers to the invention of the facts of exposition, catastrophe, presupposi- tions, and main action; the form of the comic plot, apparently, is assumed to be approximately that of the tragic plot, and the labels are applicable to both types. 1 Modern criticism, however, does not limit itself to a statement that the coherence, in a very broad sense, of later comedy is largely i Jttiij* 5e raOr' ovk &ttiv, ciWa irhvra Set eipelv, 6v6)w.Ta Kaiva, < >KaireiTa ra fiiwKTjp.ki'a irpdrtpov, to vvv irapbvTa, tt\v KaToaTpwjriiv, tt\v eUr0o\riJ>. av le -ri Tobro>v ■KapaXltrfj Xpip.T)S TM ?} QtlSwV TIS, kKOVplTTtTal.' IIijXcI Si ravr' i^eari Kai Tebupip mieCv. [Athen. 222 A, frag. 191, Kock] The contrast between Chremes and Pheidon, on the one hand, and Peleus and Teucer, on the other, seems to make certain an allusion to a comedy of manners, not to mythological travesty. The complications, the epitasis of Donatus on Terence, are covered, if at all, only in ri. w irapovra. It is quite possible that Antiphanes is referring mainly to exposition and solution. Ancient literary criticism of comedy. The Antecedents of Hellenistic Comedy 111 effected under the influence of the organic structure of Greek tragedy. It undertakes to establish a more specific structural relation between the two types. 1 The Latin plays reveal in the text conditions that point to the possibility of a "vacant stage" at intervals in the production of a given play; taking some but not all of these possibly "vacant stages" to be indications of real and essential pauses in the action, modern critics posit a division of the Latin plays into chapters of action which in Roman comedy is supposed to be an obscured reproduction of more clearly marked act division in the Greek originals; this act division in the Greek originals is itself supposed to be the result of a development in which tragedy plays a dominant part. For later Greek comedy seems on occasion, if not always, to have separated chapters of action from one another by an inorganic intermezzo chorus, or interlude scenes, or flute music — all of which might easily be substitutes for a relatively organic inter-act chorus such as, in Greek tragedy, regularly divides, or connects, the six or seven smaller chapters of action which constitute the play. The "vacant stages," therefore, of the Latin plays become a final issue in the development from a choral drama in which the chorus is organic, through later Greek comedy in which inorganic features, largely musical and often choral, marked the end of acts, to a dramatic form in which "vacant stages" providing essential pauses in the as it issues in Euanthius and Donatus, deserves more attention than it has received; the theory of structure in these Latin comments on Terence may be patchwork in its present form, but it has remote and honorable antecedents. On /catastrophe and eisbole cf. Leo, PI. Forsch. 2 , 233, and nn. 1, 2; on katastrophe I might add the mime (vs. 16) edited by Koerte, Archiv. far Papyrusforsch., VI (1913), 1 ff., with which cf. katastole in another mime (Oxyrhynch. Pap., Ill, No. 413, vs. 95) and in the scholium on Aristoph. Peace 1204. 1 If any complete analysis of the internal structure of the Latin plays had been made, I should naturally discuss it at this point. In default of such a study and for convenience in my own exposition I take up the theory of act division; for, though this problem is a matter of external and mechanical structure from one standpoint, Leo and other critics assert that the choral songs of tragedy set off logical units, and that the act division in Roman comedy often coincides with the logical chapters of the plot, as, e.g., Act I, Exposition; II-III, Complication; IV-V, Solution. This asser- tion, so far as Greek tragedy is concerned, is vigorously contested by Holzapfel, Kennt die griech. Tragddie eine Akteinteilung t (Giessen, 1914), who convinces himself that choral stasima are not at all regularly the boundaries of logical chapters, although tragedy does provide "bestimmte Richtlinien fur das Entstehen von ftinf Akten" (p. 96). I have accepted, however, Leo's assumptions in the argument above without raising the question whether or not the so-called acts in tragedy or comedy are logical units; it seems proper to meet Leo on his own ground. 112 Henry W. Pkescott action supplant the interludes that in the earlier forms kept the scenic background constantly occupied. 1 For brevity, I may state somewhat categorically the generally accepted facts, so far as I can discover them in the tangle of modern discussion: (1) Sophoclean and Euripidean tragedies fall into six or seven chapters of action set off by choral songs. (2) The Eccle- siazusae and Plutus of Aristophanes are susceptible of division into six or seven chapters; Aristophanes is supposed by some scholars to have written for these later plays choral interludes, many of which have not survived. (3) Hellenistic theory, perhaps derived from contemporary practice, divided tragedy into five acts; the practice is perhaps reflected in Senecan tragedy. (4) There is no evidence that Hellenistic comedy operated regularly with a theory of five acts, though the Epitrepontes of Menander seems, in its present fragmentary condition, to have indications of at least four acts set off by the label chorou, and the fcoraos-chorus is here and elsewhere in New comedy a distinguishing mark of division into mere. (5) It is evident that Varro and others, probably under the influence of Hellenistic theory and method, attempted with difficulty to divide the plays of Terence into five acts, and sixteenth-century editors of Plautus somewhat violently followed a similar procedure in their texts of the poet. That either Plautus or Terence consciously organized his plays into any definite number of acts is made unlikely by the known facts of Varronian act division and by the present condition of the texts, but either or both may, distinctly or obscurely, reflect act division in their Greek originals. (6) In Leo's attempt to discriminate mere in the Roman plays, using "vacant stages" and other criteria, the number of such acts varies from a minimum of three to a maximum of seven : about one-third of the total number of plays have five acts, the four-act and six-act plays are almost as numerous as the five-act plays, and divisions into three and seven acts are represented each by several plays. 2 1 In Leo's view plots of intrigue force the organic chorus out of the comedy of manners (Der Monolog, 39, 41), and ultimately the inorganic chorus is replaced by flute music or by spoken interlude (PI. Forschfi, 227, n. 3). 2 For the facts in this paragraph and further details cf . Leo, PI. Forsch. 2 , 226 ff . ; Der Monolog, 49 ff.; Legrand, Daos, 464 ff. On the fragment of the Epitrepontes, which adds a new chorou to the play, cf. Oxyrhynch. Pap., X (1914), 88 ff. For a brief summary and critique cf. Conrad, The Technique of Continuous Action in Roman Comedy (1915), 1 ff., ably reviewed by Flickinger, Class. Weekly, X (1916-17), 147 ff. The Antecedents of Hellenistic Comedy 113 This array of facts, it seems to me, is far from satisfactory as a support for the view that Hellenistic comedy owes its structural organization to tragedy. The relation between tragedy and the two later plays of Aristophanes may or may not be significant, but how or why in the Hellenistic period a five-act theory or practice developed in tragedy is unknown, and that a five-act division, or any other uniform act division, prevailed in Hellenistic comedy is not sub- stantiated by the evidence. In tragedy the chorus is the germ of the dramatic form, and as such is an inalienable organic element, which, with an occasional exception such as Agathon's embolima, only slowly acquires a detachable inorganic character. In comedy the chorus, though relatively organic in the first part of an Aristophanic play, becomes generally inorganic in the second part, in which often topical songs set off episodic dialogue; and the somewhat dubious early history of the type provides for a chorus only as an alien element. In brief, for the broad characteristics of the Menandrian komos as an inorganic element (also, of course, as composed of drunken revelers primarily) every preparation is made in earlier comedy; tragedy, on the other hand, offers inherent obstacles to such a development. It is true, however, that early comedy, as we now know it, does not furnish a structure in which logically connected chapters of dialogue are consistently set off by choral songs; this structure, now vaguely indicated in the Epitrepontes, finds a better background in fifth- century tragedy than in any known form of earlier comedy. We might easily admit the influence of tragedy in this matter if we were not troubled by the thought that in non-scurrilous comedy of the fifth century the chorus, if it continued to be employed, might have affected the structure of the plays and established a form which we may describe as resembling the present text of the Plutus, but with choral interludes replacing the label chorou in the present text of that play. This form need not have been so directly due to the influence of tragedy, but may have arisen as a compromise between non-choral Sicilian comedy and choral scurrilous comedy. In any case it is well to remember that, however a chorus may find its way into the drama at the start, once there it is very quickly made to perform desirable economic functions; the economic necessity of working with a limited number of actors and the artistic regard for a plausible representa- tion, however rough, of the lapse of time are neatly satisfied by the 114 Henry W. Prescott choral interlude; and in a non-choral drama the same objects are obtained by interlude music, by stationary scenes, or by substantial pauses. One would suppose, however, if the chorus or any kind of interlude is so important for economic purposes, that such interludes would for a considerable period in the development of drama appear whenever the dramatist needed to cover time for off-stage action, or for change of r61es, or both, and that therefore the logical unity of a chapter of action between two interludes would not be a primary consideration. It is of course likely that a new phase of action will begin after an interlude, and in course of time a conscious regard for symmetry may lead to the demarcation of logical units by interludes; and ultimately such logical chapters may be fixed in number. There is no evidence that they did become so fixed in later Greek or Roman comedy, but only that a varying number of chapters is set off by various sorts of interludes. In the Latin plays, if one is not blinded by the Euripidean theory, the visible facts are, first and primarily, that the structure in general points to a concern in the Roman theater for continuous action rather than for action interrupted by substantial pauses, least of all by any regularly recurring number of pauses in individual plays; 1 secondly, that there are in some plays conditions which, obscurely or distinctly, suggest a division into mere in the Greek originals. May I illustrate from the Persa my own attitude toward "vacant stages" and consequent act division, so far as Roman productions are concerned ? There are six possible vacant stages, at 52, 167, 250, 328, 399, and 752. At 53 ff . Saturio's monologue fills the interval of Toxilus' absence; in other words, it performs the same function as the vacant stage posited at 52, with the added and, of course, essential function of introducing us to the character of the parasite. At 168 ff. Sophoclidisca's patter-talk fills the interval of Toxilus' absence (167-83), again precisely what a substantial pause at 167 would have accomplished; why duplicate the devices for filling time intervals? 2 At 250 Sagaristio's monody similarly fills the interval 1 For the details of an argument along these lines cf . Conrad, op. tit. 2 An argument that, for example, more time is needed between 167 and 183 than is provided by the text of 167-82, and that therefore a substantial pause at 167 is required in addition to 167-82, is made difficult by the general consideration of time intervals in comedy such as Conrad sketches (op. tit., 19-34). The Antecedents of Hellenistic Comedy 115 between the departure and return of Paegnium and links two chapters of the action. At 752, just as 738-52 have made it possible for Toxi- lus departing at 737 to return at 753, so 753-76 are arranged to allow Dordalus, making his exit at 752, to return in 777; in brief, the action around the supposed vacant stage is obviously so interlocked as to serve the same economic purpose that a substantial pause in the action at 752 would adequately meet; accordingly the substantial pause becomes quite unlikely. With regard to 328 and 399 the case is different, and taken by themselves these places admit pauses so far as the text is concerned, but (1) if the other four supposed pauses are rightly eliminated it is not likely that these two places, only seventy lines apart, mark substantial breaks in otherwise continuous action; (2) a pause at 399 breaks the action at a point at which rapid action in the execution of the intrigue is highly desirable; (3) if my suggestions in CP, XI, 129, n. 2 have any validity, the distribution of r61es might point to 306-28 as devised, in part, to provide for Sophoclidisca's assuming the r61e of the parasite at 329, a condition which would make unlikely a pause at 328. 1 Now if we turn from the Latin play to the Greek original and ask ourselves whether any or all of the six possible pauses in the Latin text of the Persa were either real pauses or musical interludes of some sort in the Greek text, we face a very difficult question. We observe that the Latin text does not, implicitly or explicitly, suggest the existence in the Greek original of an inorganic chorus. And the same arguments against flute music would apply to the Greek original (if its text was essentially the same as the Latin text) as we have applied to four of the supposed vacant stages of the Latin copy. On the other hand, if the Greek text was essentially different from the Latin text, and if interludes other than monologue and monody took the place of the parasite's monologue, of Sophoclidisca's talk, Saga- ristio's monody, etc., we have difficulty in imagining just how the Greek play could have been constructed, and we also have to admit an extraordinary, not to say incredible, originality on the part of 1 That is, if there were a substantial pause at 328, this pause would supply the time needed for change of rdles, and the present condition of the text, as regards 306-28, would not be so easily explained. But of course I do not contend that the distribution of rdles in this play is so certain as to lend any great weight to this point. 116 Henry W. Prescott Plautus. 1 I leave to partisans of act division the issue; for myself I seriously question whether the Greek original of the Persa in these large structural features was essentially different from the present Latin text. Granting this, I observe with perfect equanimity that the Heautontimorumenos 2 contains evidence that an inorganic chorus operated in the Greek original; some of the vacant stages in the Latin text very distinctly point to interludes in the Greek performance such as we seem to have indicated in the Epitrepontes. And this diversity, represented in two plays, I feel perfectly free to extend indefinitely, not being hampered by any theory of exclusive or large dependence upon Greek tragedy, which inclines modern critics to put Hellenistic comedy in a strait-jacket of uniformity and regularity. 8 VII The discussion of vacant stages and of act division is much affected by the view that these and other aspects of Hellenistic and Roman comedy are the issues of a development from choral to non- choral drama. This development is suggested by many visible conditions in the texts of Old and New comedy and is explicitly stated in ancient theory, which describes Old comedy as choral and later comedy as at first removing the chorus but leaving a place for it, and then not even leaving a place for it. The last two periods of development in ancient theory are represented respectively by 1 It may be observed that there are no monologues before 52, 167, and 250 (a very- brief one before 167). This condition suggests that the solo speeches and songs at 52 ff., 167 ff., and 250 ff. are surrogates in a non-choral drama of a chorus in choral drama, in so far as they fill intervals of time primarily, though not exclusively, as does a chorus. Why may they not have performed this function in the Greek original ? 2 The conditions are particularly good at 409, where a night intervenes; at 748, where the ancillae may pass across the stage; and at 873, where the old men re-enter, having just left the stage at 872. At 229 I see no clear evidence of a break in the action; nor am I fully convinced by the arguments of Skutsch and Flickinger regarding the condition of the Greek original at 170. 3 The technique which I discern in the Greek original of the Persa is roughly analogous to admittedly Greek technique in other Roman plays in which interlude scenes, spoken or sung, are found, e.g., Captivi 460-98, 909-21, Cure. 462-86, Most. 313-47 (cf. Leo, Der Monolog, 59 and n. 2, PL Forsch.*, 227, n. 3) . Leo's contention that such spoken and sung interludes are substituted in the Greek originals for the chorus only relatively late and in the period of the technitae, I should meet with the question why they might not have appeared at any time in a non-choral drama. The Antecedents of Hellenistic Comedy 117 Menander and by the Latin poets; and the Latin poets, by not even leaving a place for the chorus, made difficult a division into the five acts which in the Greek originals were clearly distinguished by choral passages or by the label chorou. 1 That Menandrian comedy often justified such a statement of the case I see no reason to question. But ancient theory, as I have elsewhere indicated (CP, XII, 409), seems to be operating with a selected mass of material; when it speaks of Old comedy it betrays no knowledge of Crates and Phere- crates; when it discusses New comedy it often, as above, concentrates upon Menander. We may concede, however, the truth and value of the broad generalization in ancient theory without closing our minds to other facts. A non-choral type of drama has problems in common with choral drama but must meet them without a chorus. Such problems, for example, are presented by a limited number of actors, by the necessity of covering plausibly time intervals, and by peculiarities of the stage setting. Time for off-stage action and for change of r61es is easily provided by a chorus, whether organic or inorganic; non-choral drama is driven to a variety of substitutes for the chorus — to lyrical intermezzos by single actors or small groups of actors, to instrumental music, to dances, to monologues, or to dialogue scenes that may not always promote the action. It requires not a little skill to bridge gaps with scenes, whether spoken or sung, which are inseparable organic units and are not too manifestly mere bridges. With these considerations in mind we may better appreciate the most striking feature of the technique of New comedy. The difference between my own views and the tendencies of modern criticism may be illustrated by a brief criticism of Leo's theory of the monologue. 2 To understand his argument we must outline the results of his study, which, by its scope, by the thorough marshaling of material, by the nice discrimination of stylistic qual- ities, and by the historical perspective of the investigator, excites the greatest admiration and doubtless carries conviction. Racial psy- chology prepares us for an extensive use of solo speeches in Greek liter- ature. This tendency of the race is definitely limited in fifth-century 1 Euanthius De fabula iii. 1 (Wessner, I, 18) ; for further details cf. Conrad, op. cit., 8 ff., and footnotes. 2 For brevity, following Leo, I use "monologue" to cover solo speech and solo song; nor do I always differentiate soliloquy in the narrow sense. 118 Henry W. Prescott drama by the presence of a chorus. Only before the entrance of the chorus is genuine solo speech available. To this limitation set by a chorus Aeschylus and Sophocles in the main submit. Euripides, however, strains against the barrier of the chorus. His interest in solo speech led to a steady development toward a detachable prologue in the only part of the play in which he was free from the handicap of a chorus. Within the play, between the entrance and the exit songs of the chorus, a similar progress appears toward the increasing use of quasi-monologues — the prayer monologue, the address to the elements and inanimate surroundings that gradually reverts to actors or chorus, and pathetic speech that disregards the presence of chorus and actors; rarely too he removes the barrier to solo speech and, withdrawing the chorus, as, for example, in the Helena, finds expression in more nearly genuine solo speech. The quasi- monologues in the presence of the chorus Leo finds most frequently just after a choral song and at the beginning of a meros; in a relatively few cases they appear just before a choral song and at the end of a meros. The goal toward which Euripides was tending, hampered by the chorus, is clearly indicated in the Helena, a play which in so many other features of form and content anticipates later comedy. In this play Euripides reveals what he would have done without a chorus; here the mere, or acts, of the drama are bracketed between monologues with remarkable regularity. The immediate issues of this technique Leo sees in Roman comedy. The Euripidean prologue is firmly established in many plays of Plautus. The monologue, now that there is no chorus, is freely extended within the plays of Plautus and Terence, and it brackets with some regularity in many plays those units of action which Leo discriminates as mere. 1 My objection to Leo's inferences from the facts is that a sig- nificance is attached to many phenomena which they will not bear. So far as the position of the monologue is concerned, it is clear that (apart from "asides," with which Leo is not primarily occupied) the monologue as a solo speech must appear at the beginning or at the end of units of the action; at these points, only, the stage is cleared of other characters, and solo speech is possible; under any other con- ditions a solo speech must be delivered over the heads of other actors 1 For a brief resume of his argument of. Der Monolog, 53. The Antecedents op Hellenistic Comedy 119 or the chorus. In Euripidean tragedy the chorus is usually present, and the dramatist can best introduce his surrogates of the monologue only when the scene of action is relatively clear, that is, just before or after a choral song. 1 In non-choral drama a vast majority of mono- logues must appear just before the arrival or after the departure of other characters. In brief, in each type of drama the position of the monologues or quasi-monologues is largely inevitable, and it is accordingly unsafe to infer from the position of solo speeches that one type of drama has influenced the other type. The most that may be said is that Euripidean tragedy (or later Euripidean tragedy) and New comedy (at least Philemon and possibly Diphilus, according to Leo) prefer to begin new phases of the action with solo speech rather than with dialogue and much less regularly to end such chapters of the action similarly. Now this fact, just stated, may be significant and may repay careful study, but so far as Leo's main thesis is concerned, viz., that the quasi-monologues in Euripides, limited in quantity and variety, are opening the way toward the vast number of monologues in comedy, most of which are entirely different in content from their supposed Euripidean forbears, and further, that a bracketing of acts in New comedy results from Euripidean practice in this regard, we must observe, not only that the position of the monologue is an unsafe criterion and that the qualitative and quantitative differences between the two types are remarkable, but that the regularity of act structure posited by Leo for New comedy is not established by the evidence. Leo's statements of fact are full and frank, but naturally he does not throw into bold relief the obstacles to his theory. With some measure of success he finds in the Latin plays (only three) adapted from originals by Philemon the bracketing of mere by monologues. 2 Of Menander's technique he can get no clear idea because, as he asserts, so many of Menander's originals are contaminated in the Roman copies'; and in trying to account for contradictory conditions within the group of contaminated plays Leo displays an almost 1 These somewhat obvious facts are sensibly stated by Legrand, Daos, 490. 2 Der Monolog, 49-53. 3 Ibid., 55 ft. 120 Henry W. Prescott acrobatic versatility. 1 Nor is Diphilus' practice easily determined from the two Latin plays, one of which is contaminated, that come from his hand. Of ten plays not traceable to any of these three playwrights Leo finds his norm of act structure fairly well established in all but three, the Epidicus, the Curculio, and the Asinaria? Even this statement of Leo's makes a rather weak case for any dominant Euripidean influence. Without stressing statistics 3 one may fairly describe the situation in the following terms: Not a single Latin play has all its acts bracketed by monologues; ten plays, only two of which are contaminated, have absolutely no acts bracketed by monologues; eight plays alone contribute the slightest support to Leo's theory, so far as they have a reasonable percentage of acts bracketed by mono- logues (and to be quite fair I have called a little less or more than half a reasonable percentage) ; the other eight plays lie between the two extremes. If Leo contends that it is not fair to rest his case on bracketing, but that we should consider, apart from the bracketing, the proportion of acts that either begin or end with monologues, the figures are these: There are 130 opportunities to begin acts with monologues, of which the Latin plays accept 78; there are 104* opportunities to end acts with monologues, of which 31 are accepted. In other words, more than half the acts begin with monologues, and less than a third end with monologues. Or finally, not to neglect any angle, two-thirds of all the entrance monologues of Roman 1 Thus, for example, the Casina, from the Greek of Diphilus, does not accord with Leo's expectation of acts bracketed by monologues; the lindens, from the same Greek author, does accord; Leo (ibid., 54) is confirmed in his view that the Casina is contami- nated, and he sees in that play Plautine technique. The Andria, though rich in mono- logues, has no bracketing of acts; Leo (ibid., 57) remarks that Menander's composition has disappeared in the process of contamination, and that Terence's technique is that of the Casina. The Stichus, Poenulus, Pseudolus, and Miles gloriosus, on the other hand, are fairly regular in the bracketing of acts; Leo concludes (ibid., 56, 60-61) that Plautus has observed and followed the technique of his Greek originals! Obvi- ously, if one accepts Leo's theory of contamination and of the monologue, these are the only possible conclusions, but does such versatility in meeting contradictory conditions in supposedly contaminated plays stimulate confidence in theories either of contamination or of the monologue ? 2 Ibid., 59 ff . 3 The figures that follow are based on Leo's own interpretations, though he furnishes no statistics. 4 The difference between 130 and 104 is due to the fact that, conventionally, the last act of a Roman play usually ends with dialogue in trochaic septenarii, so that a monologue at the end of the play and of the last act is practically impossible. The Antecedents of Hellenistic Comedy 121 comedy (not including the Zutrittsmonolog) stand at the beginning of acts; slightly more than two-thirds of the exit monologues stand at the end of acts. In my opinion there are hardly more than two significant facts in the situation: first, as we should expect, mono- logues stand at the beginning or end of smaller or larger units of action, and in so doing must appear often at the beginning or end of Leo's acts; secondly, there is a notable predominance of entrance monologues, indicating a distinct preference for solo speech or song over dialogue in the technique of entrance; if one includes the Zutrittsmonolog and Eintrittsmonolog under the general term of entrance monologue, 60 per cent of the monologues of Roman comedy are entrance speeches, 20 per cent are exit monologues, and 20 per cent are link monologues. It is, however, more illuminating to observe the variations in practice in individual plays. For here we see, what I am most eager to establish in opposition to current opinion, the absolute negation of any uniform procedure, and the consequent weakness of a view that Euripidean tragedy exerted a determining influence upon the form of comedy. Leo himself, on coming to the two Latin plays from the hand of Apollodorus, the Phormio and Hecyra, immediately recognizes a novel and individual technique; the Phormio, for example, has twelve monologues and five acts; but only one of the dozen solo speeches stands at the beginning or end of an act, and two-thirds of them are link monologues. The Captivi, he has to admit, only seemingly supports his theory; for two of its monologues are interlude scenes, and as such reveal another novel type of structure only partially paralleled by the choragus scene of the Curculio; that is, here clearly the monologue does not follow a vacant stage but occupies a stage which would otherwise be vacant; in other words, it performs one of the main functions of a chorus. Beyond these clear marks of variety and individuality he equally clear evidences of divergence from any norm in other plays. What could be more suggestive than the contrast between the Aulularia and the Asinaria f The former is supposed by Leo to be Menandrian and is innocent of contamination; it has twenty-two monologues, an unusually large number, and four acts; yet of this large number of solo speeches only one stands at the beginning of an act, three at the ends of acts, and 122 Henry W. Prescott no act is bracketed; and all this in spite of the fact that there are nine entrance and seven exit speeches out of the twenty-two. On the other hand, observe the Asinaria, from the Greek of an obscure poet, Demophilus; it has only six monologues, the smallest number of all the plays, and all six are used in the first half of the play, one at the beginning, two at the ends of acts. Possibly the plots of these two plays are peculiar and the structure correspondingly peculiar; but are we likely to appreciate properly the various theories of act division, of monologue, of Euripidean influence, until we consider how the plot and various other factors affect structure? Between the two extremes presented by these two plays the other comedies offer other interesting vagaries, into which I need not go at present. In this discussion of the monologue I have necessarily accepted, for descriptive and argumentative purposes, the theory of vacant stages and of act division, although in the previous sections of the paper I have attacked the validity of the act theory, and of the vacant stage in Roman comedy as a criterion of division into acts. Perhaps I should state now my general attitude toward Leo's theories of the vacant stage, monologue, and act division. The broad impli- cation in his discussion seems to me to be that a rather regular sequence of exit monologue, choral song, entrance monologue in choral drama (and specially in Euripides) results in Roman comedy in a fairly uniform sequence of exit monologue, vacant stage, entrance monologue. 1 Now I am perfectly willing to admit that the r61e of the Menandrian fcomos-chorus makes it likely that a Roman poet, finding such a chorus in his Greek original, would substitute for it a vacant stage, and monologues might often appear on either side of the fcomos-chorus and of the subsequent vacant stage. What I doubt is whether this Menandrian technique was consistently employed by Menander or by other Hellenistic poets, and whether Euripidean influence is a factor to be reckoned with when such technique appears. • This statement is not quite fair to Leo. Exit monologues in Euripides are relatively few in number, and Leo would probably stress the fact that the sequence of choral song and entrance monologue in choral drama is replaced by the sequence of vacant stage and entrance monologue in Roman comedy. It is true that entrance monologue in Roman comedy is predominant, but from my standpoint the vacant stage before it is mere assumption in most cases. The sequence of exit monologue, vacant stage, and entrance monologue in Roman comedy occurs about 35 times out of a possible 104; eight plays have no examples of this sequence. The Antecedents of Hellenistic Comedy 123 In my own mind I leave room for a further possibility that, much oftener than Leo admits (in the Captivi and Curculio), a monologue is itself a substitute for the chorus of choral drama, that it bridges gaps rather than follows a gap, that it promotes continuity of action even in the Greek original, as it does in my view, for example, in the Persa as a Roman production. 1 From this standpoint possible vacant stages in Roman comedy are not very regularly substantial pauses, and monologues are sometimes surrogates of the vacant stage as well as of the chorus. So far as Euripidean influence is concerned I see nothing in the evidence that conflicts with the view that, given racial psychology which prompts soliloquy, and granting the dramatic convenience of the monologue as an artifice in facilitating structure, 2 the monologue is bound to assert itself in comedy, without any Euripidean influence, as soon as the chorus is removed; this begins to appear at once in the Ecclesiazusae and Plutus, and if the Helena also illustrates it I see only the parallel development which I should expect in the two dramatic types. 3 Euripidean influence is certainly not manifest in the spirit and general content of the comic monologue, and if its formal features are due to the tragic poet the mold has been usually filled with a content that comes either from the resources of Old comedy or from the immediate dramatic necessities of the New comedy of intrigue. The Euripidean monologue is limited in the main to prayers and 1 Leo limits the technique to the passages referred to above, p. 116, n. 3. Other passages which are chorartig (Der Monolog, 68, PI. Forsch*. 240, n. 1) in his opinion are of a different sort, being mainly Lauscherscenen. 2 Cf. Arnold, The Soliloquies of Shakespeare (New York, 1911), 81, who indulges in the paradox that the structural monologues opening, closing, and linking chapters of action are artificial speeches used to avoid the appearance of artifice. Similarly Roessler, The Soliloquy in German Drama (New York, 1915), 17, regards the structural monologue as a lubricant in the wheelwork of the drama. 8 The point will be raised that Aeschylus and Sophocles do not use the Euripidean surrogates of the monologue. In this, as in many other respects, Euripides and comedy are more or less alike, while the older tragedians differ. Modern critics hastily use this situation to establish the influence of Euripides upon comedy. But who knows, if there is any influence exerted at all, whether or not comedy as early as Epicharmus or as late as Aristophanes influenced Euripides? Euripides and Old comedy have much in common: informality, direct appeal to the people, colloquial style, indifference to sophisticated art; Aristophanes criticizes Euripides because these and other features are out of place in tragedy; it would only be a pleasant irony if the tragic poet, from unconscious sympathy or conscious imitation, often approximated the style of comedy. 124 Henry W. Peescott addresses to inanimate surroundings and to occasional pathetic speeches over the heads of actors and chorus. Of the huge number of comic monologues no general description is possible, but the commonest types are narrative monologues outlining past, present, and future action, and solo speeches on general aspects of social life. 1 The former result largely from the dramatist's obligation to cover offstage action or to make his plot intelligible; the latter, though occasionally touching Euripidean themes, 2 are quite as much in the spirit of the Aristophanic parabasis. Both types, and monologues in general in comedy, are very often explicit or implicit addresses to the audience 3 and as such reflect the informality of Old comedy; the speeches to the chorus and to the audience in Aristophanes supply all the needed literary background for the manner of delivery and for some of the material of the comic monologue of the next centuries. 4 VIII In one type of expository monologue, however, modern critics seem to have unassailable evidence of the closest interrelation between Hellenistic comedy and Euripides. The Plautine prologue that nar- rates the plot in a detachable speech to the audience delivered by a divinity, or a character in the play, or a "prologus," is generally admitted to reproduce all the essential features of the Euripidean prologue. This evidence I have no desire to minimize, but I may properly indicate by a few brief comments that the antecedents of the Plautine prologue are mixed rather than simple, as is so often the case with phenomena in which modern criticism stresses heav- ily the Euripidean features. The prologue is only one form of exposition, or only part of the exposition. At the outset I find it significant that another type of exposition, in which a dialogue between master and slave opens the play, and the master in response to urgent questions discloses facts of interest to the audience, is admitted by the chief essayist on the 1 For examples, ef . Leo, Der Monolog, 72, nn. 13 and 14. 2 Cf ., e.g., Leo, PI. Forsch?, 119; for the philosophizing as such cf. CP, XIII (1918), 134-37. 3 Leo, Der Monolog, 80; Sehaffner, De aversum loquendi ratione (Giessen, 1911), 18. * Leo, Der Monolog, 79 ft., Geschichte d. rSm. Lit., I, 107, 109, n. 1. The Antecedents of Hellenistic Comedy 125 prologues of Greek comedy not only to have distinctly mixed ante- cedents but to owe its origin to comedy rather than to tragedy; 1 and this too in spite of the closest resemblance in details of phraseology as well as of general situation between such dialogue expositions in the Roman plays and the corresponding expositions of Euripidean tragedy: ". . . . videntur mihi talia initia ut Thesmophoriazusarum Pluti Iphigeniae Aul. Pseudoli Curculionis primum ficta esse a poetis comicis, inde autem manasse et per tragoediam et per mediam novam- que comoediam." 2 Without intending at all to subscribe to any theory of origins in this matter, 3 I quote this statement of Frantz simply to suggest that in the triangular relation which is often ap- parent between Aristophanes, Euripides, and New comedy one must be open-minded to the possibility that early comedy rather than Euripides is the initiating force, and that Euripidean influence is only one of many strands in the complicated phenomenon of later comedy. It is this same triangular relation that confronts a student of the prologue as a detachable speech to the audience, if he is not biased by preconceptions of Euripides' influence upon later comedy. A discriminating critic like Leo 4 may successfully trace in the Euripidean prologues a development from a speech in which the expositor care- fully accounts in the prologue for his appearance, justifies the solilo- quy form of his address, and in general satisfies all the demands of a modern sophisticated critic, to a negligent and relatively inartistic prologue in which the speaker seems to be almost impersonal, disregards motivation, external or internal, and is conscious of the audience, though he does not directly appeal to it. 6 And the 1 Frantz, De comoediae Alt. prologis (1891), 21 ff. He is quite convinced, however, that the prologue as a detachable expository speech is thoroughly Euripidean (ibid., 30 ff., 40, 45, 49). 2 Ibid., 28. 3 In this small matter I should probably not espouse any theory of origins or influence but content myself with the observation that comic and tragic dramatists, facing similar problems of exposition, solve the difficulties in similar simple ways. The modern playwright who opens his play with a dialogue between the butler and the maid need not have read ancient drama or contemporary drama; such devices are quickly conventionalized, of course, and become traditional, but they are weak props for any thoroughgoing theory of origins or influence. * Der Monolog, 14-26. 6 Explicit address to the audience in tragedy is so rare that Frantz (op. cit., 50) properly describes it as a descent to the plane of comedy. 126 Henry W. Prescott conclusion is that in this final type of Euripidean prologue "der 'prologus' der spateren Komodie ist . . . . potentiell vorhanden." 1 Over against this fact must be balanced the equally significant con- ditions in Aristophanes' Knights and Wasps and Peace (cf. Birds, 30 ff.), which critics cannot refer to Euripidean tragedy at all; in these plays one of two slaves, after some preliminary dialogue, turns to the audience and in frankly informal address to the spectators expounds the theme or general situation. 2 Here is a much clearer background for the inartistic comic prologue of later times; nor can one deny that the interruption, in the Plautine prologues, of the exposition of the plot by facetious remarks and serious reflections (as, for example, in the Captivi) for the benefit of the audience is quite alien to Euripides and entirely in accord with the spirit of Old comedy. It is, however, quite clear that the monologue form of detachable exposition in New comedy is more closely allied to Euripidean technique than, for example, to the monologue of Dicae- opolis at the beginning of the Acharnians; and one may easily see how the travestying of tragedies would have brought over the Euripidean monologue into Middle comedy, and how readily the same expository form would have been retained in the comedy of manners. At the same time one must admit that the extreme informality, the frankness of direct address to the spectators, the conscious exposition of the plot, are all forestalled in Aristophanic comedy. In such matters Euripides may be not an initiating force but a complacent victim to the democratic informality of early comedy. Not only as relatively inorganic solo speech is the prologue in Hellenistic and Roman comedy traced to Euripides, but in the choice of persons as speakers comedy is supposed to be following closely the tragic poet. For in Euripides the prologues are delivered by char- 1 Der Monolog, 25. 2 Leo, of course, recognizes the contribution of Old comedy in this respect (Der Monolog, 80), but his general appraisal puts all the emphasis upon the Euripidean prologue. Beyer, De scaenis .... quibus .... narrantur, non aguntur (Gottingen, 1912), 49, asserting that this Aristophanic form of exposition is primitive and was established in comedy much earlier, strangely argues that it is derived from tragedy. It may be observed that, so far as this expository address to the audience in Aristopha- nes follows preliminary dialogue, it furnishes a better background for the intercalated prologue of Plautine comedy than, I think, anything that Euripides has to offer. The Antecedents op Hellenistic Comedy 127 acters in the play or by divinities; and the extant prologues of comedy are put in the mouths of the same two types of speakers; comedy, to be sure, has added to the list the impersonal "prologus," whom modern critics dismiss as a natural final development of the inorganic prologue. 1 In this bit of cumulative evidence, however, there is a deviation from complete correspondence that might prove significant of a different history for the expository prologue. The divinities who deliver the Euripidean prologues are, almost without exception, the major gods and goddesses of the hierarchy. 2 The divine beings who serve as prologists in comedy are of a different and lower order. It is at least incautious to speak of "die direkte Abkunft" 3 (from the prologizing divinities in tragedy) of such allegorical figures as Aer, Elenchos, Agnoia, Auxilium, Luxuria, Inopia, Tuche, and Phobos, and of minor deities like Arcturus, Heros, and Lar Familiaris. The consistency of allegorical prologists in comedy is striking. One may argue, of course, that the less heroic material of New comedy naturally make unavailable as prologists such divinities as Aphrodite, Artemis, Apollo, and the like, and that allegorical figures are a natural substitute for the Euripidean pro- logists. On the other hand, there seems to be no special reason why Venus should not utter the prologue of many a comedy in which the love story is prominent, or why Neptune as well as Arcturus might not introduce the Rudens, if with consequent loss of the charming detail in the present prologue; but this does not happen, so far as we can discover from extant material. 4 That Kalligeneia, who seems to have spoken an expository monologue at the beginning of Aris- tophanes' second Thesmophoriazusae, or Dorpia, who perhaps simi- larly introduced Philyllius' Herakles, is a perfect background for the allegorical prologists of later comedy is not quite certain; these deities were probably personifications of festival days, and as such approximate the divine prologists of New comedy; they may, however, have had active r61es in the plays, and the Herakles may » Leo, PI. Forsch*, 224 ft. * The case of Thanatos in the Alcestie is hardly a real exception. 8 Leo, op. cit., 212. 4 Dionysus in the Strassburg prologue is far from certain, nor are Eros and Aphro- dite in the Ghoran papyri valid exceptions. For other possible cases of comic pro- logists cf. Leo, op cit., 212, n. 4, 128 Henry W. Prescott have been a mythological travesty. But even if Old comedy had no prologists of precisely the same type as New comedy, it should be clear that the allegorical figures of Ploutos, Opora, Theoria, Eirene, the Logos Dikaios and Logos Adikos, which issue naturally from the fantastic plots of Aristophanic comedy, suggest that the allegorical prologists of New comedy, as allegorical figures, are not primarily Euripidean at all; 1 nor should anybody overlook in this connection the rdles of Earth and Sea, of Logos and Logina, in Epicharmus. The part that Sicilian-Attic comedy and very early mythological travesty of epic story and oral legend played in this development both of allegorical figures and of the prologue is unknown, but con- servative criticism will reckon with the unknown, at least so far as to modify hasty conclusions from the known. 2 IX It would strengthen the contention of modern critics appreciably if, through careful analysis of the structure of action in New comedy and of the mainsprings of action, they had established close con- nections with Euripidean tragedy. Legrand in his Daos (p. 383), having asserted, that the rigorous unity of later comedy is due to the influence of tragedy, remarks that he will, in the course of subsequent chapters, repeatedly note that the comedies employed the same motives or adopted the same general arrangement as did the dramas of Euripides; yet in his immediately following discussion of simple and intricate plots and of "les ressorts de Taction" there is not a single reference to any Euripidean parallels. In various particu- larities of dramatic technique, however, Legrand and others do find further evidence of Euripidean influence. Some representative instances of such discussions I must briefly consider. 1 The nearest approach to such figures in Euripides is in the prelude to the second part of the Hercules furens, in which Lussa, conducted by Iris, enters the palace some- what as Inopia is escorted by Luxuria to the house of the hero in the prologue of the Trinummus. I should be quite willing to grant that Philemon might have been influenced by Euripides here, without admitting that the isolated instance in Euripides is sufficient to explain the extensive use of allegorical prologists in comedy. 2 It is pertinent to remark that the call for applause at the end of the play has a background in Aristophanes; cf . Leo, op. cit., 240 and n. 3. And it is not uninteresting to observe that Leo is mistaken (ibid., 241) in thinking that the quotation of a similar tag in Suetonius (Aug. 99) is from Middle or New comedy; is it not clearly implied to come from a mime ? The Antecedents of Hellenistic Comedy 129 A characteristic of most of these studies in the minutiae of dramatic technique is the acceptance, at the start, of the Euripidean theory; the writers then proceed to find cumulative evidence of the dependence of comedy upon tragedy in whatever detail of craftsman- ship they choose for investigation. Thus, for example, Fraenkel opens a chapter of his study with the statement: "id effectum est ut hodie paene iam pueris decantatum sit ex quinti saeculi tragoedia, Euripidea imprimis, in mediam novamque comoediam non modo varia fabularum argumenta .... sed etiam singulas sententias .... defluxisse"; 1 and Harms begins his essay on motivation: "constat novam .... comoediam potius tragoediae Euripideae quam veteris comoediae forman atque rationem secutam esse." 2 Considering the vogue of the theory that Euripides is "der wahre Begriinder der neueren attischen Komodie," one can hardly blame such writers, but the danger in starting from this theory as a demon- strated fact is obvious. Nor are the methods employed in the course of investigation as sound as they should be. Constantly one finds the writers of dissertations observing that A resembles B, and that therefore B is derived from A; that both A and B may be derived from X, or that for other reasons the resemblance of A to B does not establish any causal connection between the two, never enters into their calculations. In general, having recognized the possibility of Euripidean influence, they never stop to eliminate all other possibilities. Practically such investigations are brought to a conclusion at the point where fruitful study might well begin. I can easily sympathize, for example, with anybody who, in reading the Alcestis of Euripides, remarks 3 that "der gastfreundliche Herr, die aufopferungswillige Gattin, der treue, etwas beschrankte Diener, der bose, senile Alte" can easily be paralleled from Menander and Plautus. But if this stereotyping tendency in Euripides is a natural issue from the technique of the Marchen, and if in 1 Fraenkel, De media et nova comoedia qu. sel. (Gottingen, 1912), 53. 2 Harms, De introitu personarum in Euripidis et novae comoediae fab. (Gottingen, 1914), 1. s Howald, Untersuch. zur Technik der euripid. Trag. (Tubingen, 1914), 19. Howald does not use the resemblance to prove any interrelation, and I quote his words only to illustrate a natural and current impression of the likeness between Euripides and comedy in the matter of characters. 130 Henry W. Prescott Aristophanes I find Socrates approximating a typical philosopher rather than the real Socrates, and if Aristophanes and Doric farce already have developed, without Euripidean influence, stereotyped professional r61es, I must conclude that the degree of Euripidean influence upon New comedy in this respect is difficult to determine; certainly I cannot lay much weight on the fact that Aristophanes does not stereotype domestic roles as long as he has little occasion to use them. And I must remain open-minded to the possibility that the resemblance of Euripides to New comedy does not establish any interdependence of the one and the other. For aught I know Sicilian-Attic comedy may have had stereotyped domestic r61es before Euripides wrote tragedy. I am not denying that some comic poets learned something about character treatment from Euripides, directly or indirectly, but again the whole problem is a complex, not a simple, one. The so-called unities of time and place in drama have been studied, and various observations have been made regarding the devices used by dramatists to preserve these unities. The recent history of such studies is significant. Felsch 1 records the artifices used by Greek tragedians. Polczyk 2 follows with a study of the same problems in New comedy and notes in connection with almost every artifice that Felsch has found the same device in Greek tragedy; Polcyzk then concludes that in these respects New comedy is depend- ent upon tragedy. Almost immediately, however, Todd, in studying the unity of time in Aristophanes, avows that Old comedy uses the same devices as New comedy, a fact which Polcyzk had denied. 3 If Todd is right we are confronted with a dilemma: Did Euripides teach Aristophanes these artifices ? Or did Aristophanes, Euripides, and poets of the New comedy, facing the same problem, solve it in the same way independently of one another ? That the second of these two alternatives must be chosen seems to me likely when we are concerned with particularities of technique that are clearly due to conditions of the Greek theater, in which Euripides and poets of the Old and the New comedy produced their iBresl. Philol. Abhandl., IX (1907), Heft 4. 2 Polczyk, De unitatibus et loci el temporis in nov. com. obs. (Breslau, 1909). "Todd, Han. Stud. Class. Phil., XXIV (1915), 50 ff. The Antecedents of Hellenistic Comedt 131 plays. A rigid scenic background and an essentially outdoor setting were conditions that faced Euripides and the comic poets; resem- blances between tragedy and comedy, therefore, in artifices which manifestly result from a common interest in overcoming these and similar difficulties cannot be used to establish the dependence of comedy upon tragedy, especially when the devices are of a simple and obvious nature. So in the mass of conventions relating to the mise en scene which Legrand accumulates on pages 428-63 of his Daos, nobody should look for any evidence of the interrelation of the two literary types; nor does Legrand venture beyond the wise statement (p. 461) that the germs of these conventions, are found both in Aristophanes and in Euripides. Other critics rashly jump to con- clusions; even if Polczyk is right in denying that Aristophanes preserves unity of place by the same devices as Euripides and New comedy, it is hazardous for him to argue from the resemblance in this respect between the tragic poet and later comic poets that New comedy took over these conventions from tragedy. It is of course natural, when Aristophanes differs in his procedure, and Euripides and New comedy agree, to infer a close historical relation between tragedy and New comedy. Even this inference is unsafe if, as is the case, tragedy and later comedy have in common but quite independently of each other domestic plots and broadly emotional incidents which Aristophanes does not employ. Thus, for example, Harms 1 in his study of motivation observes that the entrance of characters upon the stage in Euripides and in New comedy is often motivated "aut dolore aut inquiete animi aut consideratione," whereas in Aristophanes such emotional and mental conditions are not generally employed to make the entrance of characters natural and inevitable; for this and other reasons Harms concludes that New comedy takes over from Euripides its devices for motivating entrance. But when such resemblances are pointed out one should first consider whether the common element may not be accounted for without any dependence of one type upon the other. Aris- tophanic comedy, in the nature of the case, does not stress the emotional side of life; Euripides and New comedy, on the contrary, are dealing with the emotional experiences of everyday people and 1 Op. Ht., 64. 132 Henry W. Peescott will naturally motivate action by elementary emotions to a very large degree without necessarily being interdependent in that respect. Harms, and others in similar studies, apparently strengthen their arguments by pointing out corroborating resemblances in details of phraseology and style. This procedure is in itself quite legitimate, but again the critics are hasty in their inferences. In the first place, some stylistic features which Euripides cultivates became common property of writers in the Hellenistic period and may appear in New comedy without any direct influence of the tragic poet. Again many details of form and turns of phrase may recur in both types of literature, because they are taken by each type, independently of the other, from the common fund of colloquial expression which Euripides, somewhat abnormally, and comedy, quite naturally, delight to use, or from some other common source. When Harms, 1 for example, discovers that the Phrygian in Orestes 1375 justifies his entrance by revealing fear in the words, "Woe's me; whither shall I flee?" and that Bromia in the Amphitruo similarly exclaims, "me miseram, quid agam nescio?" and Myrrhina in the Hecyra, "perii, quid agam ? quo me vortam ? " the resemblance in these emotional commonplaces between Euripides and New comedy moves me about as much as would the discovery that Harms and I had made the same blunder; without imitating him I am quite capable of it. Each of these details is trivial in itself, but the discussion of them so pervades the treatment of comedy in these days that I may be allowed another concrete example. FraenkeP discovers the following feature in both Euripidean tragedy and later comedy: Two interlocutors in a dialogue scene are engaged in expounding a situation or facts; one of them, A, is telling the story, but instead of setting it forth in an unbroken sequence he interrupts himself and turns to the other interlocutor, B, and says, "Do you know so-and-so?" B answers, "Of course I do," and there follows a brief conversation on this line, after which A resumes his narrative. Now this simple bit of dialogue technique Fraenkel offers as proof of the dependence of comedy upon Euripides, although he says incidentally, " sane e cottidiani sermonis consuetudine mutuatus." Naturally I wonder how he knows that i Op. tit., 29 ff. 2 Op. tit., 54 ff. The Antecedents of Hellenistic Comedy 133 Euripides and any comic poet did not independently draw upon the material of ordinary speech for this device; and I wonder too just how any dramatic poet who prefers dialogue to monologue can manage a bit of expository narrative without some such common- place device by which the other person in the scene may be drawn into the conversation. Briefly then, in these particularities of technique modern criticism stops short at the simple equation of resemblance with dependence. But to establish dependence something more must be discovered than simple devices to meet conditions, external or internal, that are common to both types of drama and result either from production in the same sort of theater with similar peculiarities of scene setting, or from the use of similar pathetic material. X The force of these modern tendencies has led us to view Roman comedy as a Kunst, either quite disregarding farcical and burlesque elements and inorganic structure, or dismissing them as Roman intrusions in the artistic fabric woven under Euripidean influence. We need feel under no obligation to demolish this theory of Euripidean influence; least of all need we set up an opposing theory. But, as often in the study of literary genesis, a confession of ignorance is a wholesome preliminary to the discovery of sound methods and of helpful results. Surely we must admit that the direct and indirect literary antecedents of Hellenistic comedy include a number of totally unknown factors. There is the transitional period of Middle comedy, represented only by fragments; there is Sicilian- Attic comedy, of whose form and content we are quite ignorant; there are, possibly, subsidiary factors, like the mime and fictitious narrative in prose, which are chiefly known to us now only as they were developed in centuries later even than the period of New comedy. Such condi- tions should promote a conservative attitude toward any theorizing. It is very tempting to seize upon' the known extant material of Euripides and Plautus and Terence and to construct a theory of dependence that disregards the unknown. Some degree of substantial dependence upon Euripides in particu- lar and tragedy in general is made probable by the cultivation of 134 Henry W. Pbescott mythological travesty in the Middle period. The general proba- bility, however, and the degree of dependence are very difficult to determine, in view of the loss of comedies from the transitional period, and must be qualified by two known facts: (1) that such mythological travesty is much earlier than the Middle period and dates back even to a time when epic and oral tradition of myth may have been the subjects of travesty; and (2) that Aristotle seems to have found in Sicilian-Attic comedy rather than in Aristophanes or Euripides the antecedents of the comedy of his own day. In addition to the indirect influence of tragedy through mythological travesty there is a more palpable and immediate impact of tragedy upon a few individual poets, notably Menander and Philemon; yet the general character and degree of such influence hardly warrants a careful critic in demanding even of Menander and Philemon a regular conformity to supposed canons of Euripidean art. And at least the current assumption that Hellenistic comedy as a whole was monoto- nously regular and uniformly artistic deserves a thorough overhauling. Mere comparison of Euripides and New comedy may lead to deceptive results. Currents of thought that are abnormal in the time of Euripides become commonplace in the next century; demo- cratic informality that sets Euripides apart from Aeschylus and Sophocles is an inherent quality of all comedy in Greece; prosaic and colloquial idioms that are idiosyncrasies in the tragic poet are the natural stock in trade of comedy; the material of later comedy is pathetic, as, independently, are the incidents of tragedy; and, finally, tragedy and comedy were produced under roughly the same external conditions. Naturally, therefore, there will be resemblances, but only after careful study may we accept them as evidence of any direct influence of tragedy upon comedy. Like many other types of literature in the Hellenistic period, comedy marks the confluence of many different streams, the crisscrossing of various earlier types, the constant fusion of contemporary realistic experience with themes and incidents conventionalized by a conservative literary tradition. A frank recognition of the complicated phenomenon would save us from the dangerous use of simple universal solvents. Our present practice, based on the Euripidean theory, is treacherously easy. We measure all the plays of Plautus and Terence by the standard of The Antecedents of Hellenistic Comedy 135 Menander's Epitrepontes 1 and assume a uniform procedure in all the Greek authors of the originals which Plautus and Terence adapted, blinding ourselves to the manifest variety in the twenty-six Latin plays. With supposed canons of Euripidean art as a basis we note the inartistic and attribute it to Roman handling, disregarding both the fact that the whole history of Greek comedy naturalizes inartistic irregularities and the likelihood that the tradition of the Latin texts through the hands of stage managers offered every opportunity for excision, substitution, and displacement. What little we know and the large amount of what we do not know should lead us to approach the higher criticism of Roman comedy with caution and in a somewhat pessimistic temper. But there is one condition that prompts a mildly optimistic outlook. Twenty-six plays constitute a considerable mass of material. Should it not be possible, disregarding all theories, to analyze these plays, placing side by side like features, discriminating the unlike, and thereby ultimately obtaining a helpful synthesis, which might lead to sounder constructive interpretation? Legrand, in his Daos, has made a notable attempt to co-ordinate some important facts, but many problems remain either untouched or, if handled at all, only blurred by the shadow of the Euripidean theory. The results would not be startling; many difficulties would remain unsolved; the neatness and despatch of recent dissection, which removes the excrescences of Roman botchwork from the sound body of Euripidean Kunst, would be wanting; but we should at least be starting from a very proper confession of ignorance instead of from a mere theory that is supported, in large part though not wholly, by various weak hypotheses. University op Chicago 1 Cf. Wilamowitz, Sitzb. der. berlin. Akad. (1911), 485.