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. PETRONIUS AND THE GREEK ROMANCE By C. W. Mendell Some years ago Professor Abbott published in Classical Phi- lology 1 a stimulating article entitled "The Origin of the Realistic Romance among the Romans." In that article he indicated many- possible sources from which Petronius may have drawn something of his tone or matter. As Abbott himself suggests, all of these are sources for various specific characteristics of Petronius rather than ancestors from which the literary type proceeded. He concludes with the statement that "so far as our present information goes, Petronius seems to have been the inventor of the realistic romance." Among the possible sources of Petronius, Abbott mentions the love romances of the Greeks which, if we could be sure that they were written in their developed form as early as the time of Petronius, would furnish a type of source different from the rest. Heinze 2 assumed for the romances an early date and developed the theory that Petronius wrote a parody of them. He finds in this way a fore- runner of the type rather than of particular characteristics. I believe that this is the right direction in which to look for the literary ancestor of Petronius and that there are more indications of this relation than Heinze makes use of; furthermore, that Abbott's enumeration of characteristics reinforces rather than controverts this view, but that Heinze's theory of the parody nature of the novel is not the correct one. I am skeptical of the propriety of calling Petronius' work a real- istic romance, if that implies, as it seems to, an attempt on the part of the author to present human life essentially as it is. 3 The term "realistic" was presumably first applied to the book because it dealt *VI (1911), 257 ff. 2 R. Heinze, "Petron und der griechische Roman," Hermes, XXXIV (1899), 494 ff. » This is the position maintained by Martin Rosenbliith in a Kiel dissertation: Beitr&ge zur Quellenkunde von Petrous Satiren (Berlin, 1909). Burger also calls Petronius' work ein echt realistischen Sittenroman; see "Der antike Roman vor Petro- nius," Hermes, XXVII (1892), 345 ff. [Classical Philologi XII, April, 1917] 158 Petbonius and the Greek Romance 159 with everyday folk, not with superhuman or heroic characters. But so did the love romances. That the hero and heroine were superla- tively beautiful and in most ways extremely noble did not of necessity remove them from the sphere of possibility. Frankly accepted impossibilities were excluded from these romances, indicating a con- scious effort at realism. But they still without apology introduced improbabilities of an extreme sort, and a succession of unusual expe- riences which in their total are quite incredible even though no par- ticular one is by itself impossible. And this is just what Petronius does. He never asks us to believe in marvels, but he sends his characters through a series of adventures which the most credulous mind could not find probable. And even so we have but a small portion of them. A parallel from another literary field may make more clear the actual position of Petronius in the scale of realism. Greek tragedy dealt with exalted characters of the heroic past; gods and goddesses and personages purely mythological. Aeschylus did not scruple to deal with impossibilities; there was good ground for Aristophanes' thrusts at his horse-cocks and goat-stags; his gods and his sea-nymphs and his winged Erinyes all trod the earth among his characters. Euripides made the drama much more realistic, and although mythical kings and princes and the dwellers on Olympus still make up the personnel, they show the emotions and actions of real people. The New Comedy took the next step. The logical successor of Euripides' tragedy, it reduced every detail to the plane of the common- place. Ordinary people with ordinary emotions are its material, and it is called, in contrast to tragedy, "the mirror of life." But it is a speculum vitae only by comparison. It depicts only one side of life and that too distorted by exaggeration, with experiences tread- ing on the heels of one another in such rapid succession and with such a persistently comic phase that they can hardly be termed, with any strictness of expression, realistic. What is true of the New Comedy in this respect is roughly true of Petronius' work. It does not show the marvels of the early tales of adventurous travel. It confines itself to what might conceivably happen, but not to any truthfully realistic depiction of life. The romance of adventure, pure and simple, presented the frankly 160 C. W. Mendell impossible with the utmost confidence. The romance of love confined itself to the realms of possibility so far as incidents go and, like Euripides, presented persons of an exalted rank, in a serious fashion, acting in a natural and human manner. Petronius reduced the characters to middle and low society and dwelt on a very different phase of their experiences. Just what that phase was and the resultant tone of the treatment will appear later. The acceptance of some such evolution of the prose romance does not imply that with the beginning of one type the preceding ceased to be written, but merely that in a general way this was the order of their first appearance and development of type. 1 This Heinze doubts. He thinks that there is no such relation between the romance of adventure and the erotic romance. They are, he holds, two totally different and unrelated types, alike only in the one point that their plots are not stationary. He seems, however, to be influenced by the fact that he places the only writer of the romance of adventure that we know, Antonius Diogenes, too late to influence Petronius, while at the same time he posits erotic romances before the time of the Roman novelist. This is too cavalier. Even with- out the evidence of Lucian in the introduction to his True History, it would be clear that the romance of adventure was a very slight variation from the professedly historic work of such logographers as Ctesias and Iamboulos. Without positive proof it is almost impos- sible to give up the generally accepted theory of development which recognizes a logical sequence from the novel of adventure to the erotic romance. The dating of Greek romances is at best a hazardous undertaking. But there is certainly no ground for putting Antonius Diogenes later than Petronius. It has been assumed that his bilingual name dis- proves the statement of Photius that he lived shortly after the time of Alexander. 2 But Livius Andronicus came to Rome in 275 b.c; Rome had come much into contact with Greek cities in such a way as to acquire Greek slaves long before that; Naples and Cumae were 1 This distinction is clearly drawn by Otmar Schissel von Fleschenberg in Ent- wickelungsgeschichte des griechischen Romanes im Alterthum (Halle, 1913) . 2 This is the position of J. S. Phillimore in his article on the romances in English Literature and the Classics (Oxford, 1912). Peteonixjs and the Gbeek Romance 161 Roman before 300; and before Alexander began his conquests Rome had conquered many a Greek in Campania. Furthermore, regardless of exact dating, his nearness to Ctesias and the other logographers makes it more than likely that he wrote before formal rhetoric had begun its sway. The erotic romances, too, must be given much earlier dates than they used to receive. Jebb stated without hesitation that Chariton was a writer of the fifth century, 1 but a papyrus of 100 a.d. with a considerable fragment of his novel deducts from this date more than two hundred years at a blow. 2 The Ninos fragment cannot have been written later than 50 a.d., and in all probability was distinctly earlier. 3 These two romances have one thing in common which seems to me significant. Chariton's novel has much more deliber- ate rhetoric than the others that we possess, and the Ninos fragments indicate that that romance was similar to Chariton's in this respect. In psychological analysis, in balanced arguments, in brilliant descrip- tions, and in dramatic scenes, Chariton is far ahead of Xenophon, for example, or Achilles Tatius. It seems therefore not improbable that, although Rohde was wrong in his actual date for Chariton, he was right in his relative dating, and that the two romances of whose dates we have some slight indication are among the latest. 4 If Chariton is two hundred years and more older than he was generally believed to be, there is no reason to think that the rest may not be too, and internal evidence strongly suggests that they mark an earlier stage of development. In consideration of the possibility of an early date for the romances, it seems not unlikely that too little weight has been given to a passage from Plautus. 5 In Men. 247, Messenio says to his 1 See A Companion to Greek Studies, Cambridge, 1906, p. 161. 2 Grenfell, Hunt, and Hogarth, Fayum Towns and Their Papyri, p. 74. "See U. Wilcken, "Ein neuer griechischer Roman," Hermes, XXVIII (1S93), 161 ff. * Rohde, Der griechische Roman und seine Vorlaufer, 2d ed., p. 521: "Nur so viel scheint eine genauere Betrachtung seines Romanes zu lehren, dass er die Romane des Iamblichus, Heliodorus und nicht am wenigsten den des Xenophon vor Augen hatte und nachbildete." A fragment of Achilles Tatius of a date not later than 300 puts another of the romances much earlier than Rohde's dating. See Oxyrhyncus Papyri, X, 135, No. 1250. » Professor Henry W. Prescott has kindly called to my attention the fact that this point is brought out by Bousset, Zeitschr. f. d. Neutestamentl. Wiss., 1904, pp. 18 ff. 162 C. W. Mendell master: "Quin nos hinc domum | redimus nisi si historiam scripturi sumus?" He has already complained of the traveling as follows: "Histros Hispanos Massilienses Hiluros | mare superum omne Graeciamque exoticam | orasque ItaUcas omnis, qua adgreditur mare | sumus circumvecti." This sounds more like material for a romance than for a history, and his conclusion indicates the further ground of similarity which led him to associate their wanderings with those of the romantic hero : " Hominem inter vivos quaeritamus mortuom; | nam invenissemus iam diu, sei viveret." There is nothing to contradict this interpretation in the other instances of the use of historia in Plautus. 1 It is obvious since the discovery of the Ninos fragments that the rhetoric which could have influenced the romance was the rhetoric taught by such men as Seneca and his predecessors, not the rhetoric of the new sophistic. The balancing of arguments and the descrip- tive chapters are exactly the sort of thing which the regular rhetorical training would have cultivated. It by no means follows that the romances were expanded rhetorical exercises. The condensed plots which Seneca and the other rhetoricians collected were much more probably summaries of longer stories which the teacher pre- sented in the form of an abstract for the pupil to practice on. The fact that many of them deal with actual historical incidents bears out this supposition and the modern case system of studying law fur- nishes a fair parallel. That the school teacher invented and the novelist borrowed is certainly a harder theory than the reverse. Parthenius, some thirty years before Christ, culling plots for Gallus to use, is a good example of the sort of thing that Seneca probably did to obtain for his classroom the material that has survived him. The interesting point for the present study is that he evidently had romances from which to draw. Another parallel from the drama may serve to clarify the under- standing of this development of type. Even in the time of Aeschy- lus the desire for the romantic element was strong. Colonization 1 Bacchides i. 2. 50; Trinummus ii. 2. 100. R. Reitzenstein, Das Marchen von Amor und Psyche bei Apuleius, pp. 32 ft. and 62 ff ., interprets historia and fabula differ- ently, but to take historiae inseruisse as "wrote at intervals while engaged on a serious work" is very hard, and the whole argument is based on the unfounded claim that Apuleius found his entire matter in Sisenna. Petbonius and the Gbeek Romance 163 had no doubt fostered it. In Aeschylus it is satisfied by the marvel- ous, the supernatural, and the strange. Such long stories as Io's of her wanderings have little reason for existing except as they cater to a public craving for romance. In the later dramatists, who catered to less naive audiences, the strange and weird drop out and the love element enters to make good the romantic loss. Euripides incurs the charge of degrading the stage by introducing women in love among his characters. Finally, in the New Comedy, the love element is more regularly the central theme, but is reduced to a lower plane and treated with much less of dignity and seriousness. So the romantic element in prose story-telling proceeds from the original travel motif with its marvels of imagination, the sort of thing that Lucian reproduces as an extravaganza in his True History, to the love story with adventure as a contributory element. And when it has run its course on the high and arid plane of pure romance, it is revived by an infusion of Roman salt by Petronius. Even in the part that rhetoric plays in these two very different fields the parallel holds roughly. The rhetoric of the Greek romance is the rhetoric of Euripides; the naturalness of Petronius is the naturalness of the New Comedy. If we could be absolutely sure what the Milesian Tales were and just what Aristides wrote, one step in the development of the romance might be cleared up. But entire agreement on this point seems impossible; I merely venture a suggestion. Ovid's phrase (Tristia ii. 413), "Iunxit Aristides Milesia crimina [or carmina] secum, Pulsus Aristides nee tamen urbe sua est," is pretty generally taken to indi- cate that Aristides formed some kind of a whole out of the Milesian stories which he found or invented. The tone of the Milesian tale was presumably erotic and piquant. 1 Now it seems to me that the second reference of Ovid to Aristides (Tristia ii. 443) implies that his book was in reality a consecutive story: "Vertit Aristidem Sisenna nee obfuit illi Historiae turpis inseruisse iocos." Fabula and not historia is the regular word for short incidental stories. Many references to Petronius and Apuleius confirm the results of 1 The best summaries of the evidence are Hans Lucas, "Zu den Milesiaca des Aristides," Philologus, LXVI (1907), 16 ff., and Otmar Sehissel von Fleschenberg, Die griechische Novelle (Halle, 1913). 164 C. W. Mendell a study of Ovid's usage on this point. 1 Historia indicates something sustained, 2 and such a meaning is demanded by the passage in order to make possible the insertion of the ioci which, as elsewhere in Ovid, are evidently erotic anecdotes. Sisenna then inserted short anec- dotes into his translation of the book of Aristides. (Possibly Ovid was misled into thinking that the original book of Aristides was a compilation, by the title of it which was very likely Milesiaca after the analogy of the Babylonica of Iamblichus and the Indica and the like of his predecessors.) The resultant book was the novel that caused scandal when found in the luggage of one of Crassus' officers after the battle of Carrhae, and it was probably in its Latin version that it was known to the Romans who refer to it. So it is not strange that the scandalous insertions came to be looked on by them as the essential part of the romance and that the name of the Greek romance that gave them a setting was misconstrued into a title for them. In view of some such possibility, Apuleius' statement about his own work in his introduction is significant: "At ego tibi sermone isto Milesio varias fabulas conseram," etc. He means that in addition to merely Latinizing the Onos of Lucius of Patrae he will insert stories of erotic flavor and it is in these that much of his originality lies. He speaks of the inserts as fabulae, with the exception noted above, as does Petronius in introducing the story of the Matron of Ephesus (110, 113). If this supposition about the Milesian Tales be true, then it is in the Latin romances only, in Sisenna, Petronius, and Apuleius, that there are short inserted anecdotes. The Greek romances have numberless episodic digressions which seem to take the reader far afield, but fabulae deliberately introduced as short stories they do not have. It looks very much as though the introduction of these were a Roman contribution to the romance. But before considering this point further it will be convenient to consider the characteristics of Petronius as presented by Abbott and to see how they bear on the suggested relation between the Greek and the Roman romance. > Cf. Petr. 39, 61, 92, 110, etc.; Apul. i. 20, 26; ii. 15, 20, etc.; Ovid Met. iv. 53; Ex ponto iii. 2. 97; Tristia iv. 10. 68, etc. 2 Cf. Quint. Inst. Orat. ii. 4. 2; in Apul. viii. 1, historia is used instead of the usual fabula and the insertion turns out to be a miniature romance. Petronitjs and the Gkeek Romance 165 Professor Abbott, in outlining the characteristics of Petronius' novel that must be considered in looking for a predecessor, notes, first, the place, Southern Italy; second, the generally low class of characters; third,, the prominence of women; fourth, the efficient motif, erotic in spite of the framework furnished by an offended deity; fifth, the baffling tone of satire; sixth, the realism; seventh, the character-drawing, and, finally, the prose-poetic form. These must be compared one by one with the characteristics of the romance. The scenes of the Greek novels range pretty freely over the known world, although one point is worth noting which I think has not been brought out. The range of scene is confined to the old Greek world, not to the expanded Roman world. Persia, Asia Minor, Egypt, Southern Italy, Sicily form the stage on which the characters of romance move. If the novels were a late invention, if they did not, as I believe they do, go back to Hellenistic times, this would be diffi- cult to explain, but it is perfectly natural in the case of a type of literature firmly fixed before Rome had widened the bounds of the world. The early romance of adventure had treated in splendidly cavalier style Scythia and Thule, the lower world and the moon, until the tendency toward realism led the romancer to limit himself to the world his readers knew. But to return to the point, the scenes cover always a considerable range. Southern Italy and Sicily figure in Xenophon of Ephesus and also in Chariton, who is the representative of the romance who comes nearest to Petronius. The scene of the Roman novel in the fragment we possess is Southern Italy; there is indication that Massilia, too, figured in the story, and the haphazard wandering of the characters leaves little room for doubt that the scenes were even more varied. The characters of the romances are mixed. From the great king of Persia to a humble herdsman, all stages of society are represented. The hero and heroine, ordinarily half-way between these extremes, are, in the greatest variety of ways, brought into contact with prince and pauper, pirate and pander. In Petronius there are no real potentates. But, granted the criteria by which his middle-class folk gauge position, the range is like that of the Greek. From Habinnas and Trimalchio to the fisherman who picks up the heroes after the wreck, we have the counterparts of the romantic satrap 166 C. W. Mendell and vassal. The general type of character must be noted (it is dis- tinctly low by comparison), but the range of character is just as important, and in that Chariton and Petronius are alike. In the matter of the prominence of women, I can find no reason to draw any distinction between Petronius and the romance. In Chariton, Callirhoe, Plangon, and Stateira are very prominent and sharply differentiated, and every romance furnishes its quota of well- drawn women. Manto, the Potiphar's wife of Xenophon, his ugly Kuno, Melite, the scheming widow of, Achilles Tatius, the splendid farmer's daughter of Iamblichus— these can hardly be said to play a less important part than men. In motif again there is little distinction beyond that of tone. The efficient motif of the romances is erotic just as much as is that of Petronius. In the romance it is circumstance that drives on the lovers through a mad succession of experiences even though these circumstances are sometimes motivated by the anger of a god offended by the obstinacy or arrogance of the lovers. Petronius, too, gives realism by making the force of circumstances govern the plot, but behind circumstances, he reminds us several times, is the anger of the offended Priapus. Like the low caste of his characters this choice of offended deity must be remembered as significant. The baffling tone of satire is peculiar to Petronius. The romances are beyond all else ingenuous. The realism I have already discussed. The trend away from marvels had gone as far in Chariton as in Petronius. The only difference lies in the class of people treated: in keeping with his middle-class people, Petronius' adventures and melodramatic situations are middle class. I am inclined also to detect less difference than is usually found in the matter of character- drawing. What difference there is seems to me to be a result of the individual skill of a particular artist, not a question of type. As psychological analysis began to enter into the romance, characteri- zation began to be more prominent. It is one of the chief results of strong rhetorical influences. And Chariton in particular depicts individualities: the faithful Polycharmus, a second Pylades, the gentleman Dionysius, the oriental queen Stateira, the big-hearted countrywoman Plangon; they are all real characters though not done with the genius of Petronius. Few of the romancers seem to Petronius and the Greek Romance 167 have been men of high genius. Finally, the prose-poetic form is peculiar to Petronius. It is clear, I think, that in most of its characteristics Petronius' book is not far removed from the Greek romance. Further points of similarity might be noted. Heinze has dwelt on the framework and motivation, on the union and separation of Giton and Encol- pius, on the steady chain of misfortunes, on the constant erotic temptations of the heroes. I would add the insistence on hair- breadth escapes which are the mainstay of both the Roman and the Greek novelist in holding attention, and the prevailing willingness of the characters to give up and die when crises arise. Details, too, bear out the parallel which framework and general characteristics make so clear. To name a few: the cloak motif or recognition by means of a garment (Petr. 12 ff .) is almost identical in the use made of it with the incident of Iamblichus in which Sinonis tries to sell a cloak and is arrested for robbing a tomb. Encolpius before the pictures (Petr. 83) recalls the opening of Achilles Tatius. The argument over punishment (Petr. 107) may be paralled in any romance, perhaps best by the arguments of the pirates and the trial scene in Chariton. The shipwreck (Petr. 114, 115) might be taken bodily from Achilles Tatius or Xenophon of Ephesus. The comparison of Circe's beauty to that of a goddess (Petr. 126) suggests the familiar conceit that the Greek heroine would be taken for Aphrodite on the street. The magic potions of Greek romance are perhaps paralleled by the dis- gusting magic of Petronius (Petr. 134 ff.). With all these points of similarity which cannot be mere coin- cidences, there remain the essential differences noted in passing: the class of characters treated, the baffling tone of satire, the prose- poetic form, and I would add the insertion of anecdotes unrelated to the plot. It is time to consider the name under which the novel of Petronius has come down to us. Buecheler's Petronii Satirae is the accepted designation, but this title does not describe the book correctly if it implies conformity with one of the two types of satire distinguished by Quintilian, the Horatian and the Varronian. The hexameter form of the first is an unsurmountable obstacle, and Varro like Horace wrote short satires collected into books, not one long and 168 C. W. Mendell continuous composition divided into chapters. It is not surprising, therefore, to find that there is practically no manuscript authority for designating Petronius' book as a satire. On the other hand, it will not be difficult to see how the designation crept in during later years. The majority of the best manuscripts read saturicon, either in superscription or in colophon (so BDEFG 1 ). B, the oldest, has " Petronii Arbitri Saturicon." In the eleventh-century Paris manu- script alone appears any form of the word satura: "Petronii arbitri satirarum 1. incipit," and this is corrected in the margin by the same or a contemporary hand to "Petronii arbitri affranii Satirici lib. incip." This has the appearance of being an attempt to remedy the strange-looking form satiricon, further changed to Satyri in the Trau manuscript (A), and modern editors have been equally prone to change it to suit their purposes. Savaron in his notes on Sidonius Apollinaris 2 says: "Petronii libellus mera est satura Varroniana, ut suo loco dicitur, ipse tamen Saturicum sive Satiricum maluit inscri- bere: quo modo commentarium dicitur pro commentario libro: Apologeticus pro apologetico libro." Casaubon says 3 that it is not worth while discussing why Petronius preferred Satiricum to Satyra. As a matter of fact there is no reason to think that he did. Saturicon is in all probability the Latinization of the Greek genitive plural, and the satirarum of P, which has been largely responsible for the classi- fication of Petronius' novel as a satire, was an attempt by the copyist to correct this. The longer fragment of the Trau manuscript (A) shows that there were at least sixteen books in the novel so that /Si/3At