SOURCE: "Petronius and the Comic Romance," Classical Philology," Vol. XX, No. 1, January, 1925, pp. 31-49.

[In the following essay, Perry rejects several proposed literary forerunners of the Satyricon, contending that its more likely model was the straightforward comic narrative.]

In the present state of our knowledge, and owing to the nature of the problem itself, any attempt to account for the origin and peculiarities of Petronius' Satyricon must involve, at one point or another, the assumption of something that cannot be definitely proved. The following study is subject, of course, to these limitations. It is undertaken, however, in the belief that certain facts of ancient literary history have not hitherto received their proper evaluation in this connection, and that some advance may yet be made toward a more probable and comprehensive solution of this important problem.¹

The Satyricon, or rather what remains of it, relates, in autobiographical form, the low-life adventures of a degenerate rhetorician, Encolpius—a fellow of negative character who lives, though not very successfully, by his wits and by the arts of the parasite. Accompanied by a young favorite named Giton, whose loyalty constantly wavers, this anti-hero, Encolpius, wanders aimlessly about, constantly involved in ludicrous intrigues with low, though sometimes educated, companions and everywhere pursued, it would seem, by the wrath of an offended Priapus.² Into this general framework are introduced … such elaborate side shows as Trimalchio's dinner, the brilliant harangues on the decay of liberal arts, the long poem on the Civil War, or the story of the matron of Ephesus. The realistic portrayals of men and manners throughout combine to give us a gay, but often grotesque panorama of society unmasked and unrobed; and, to borrow a phrase from Petronius himself, "everything resounds with mimic laughter."³

Such, in brief, is the Satyricon. When we ask ourselves how such a story came to be written, and what known type, or types, of literature may have served as its chief model or forerunner, we are confronted with several interesting possibilities.⁴ The Menippean satire, the mime, the epic, the Greek erotic romance, and the Milesian tales have each apparently contributed something to the tone, or the subject-matter, or the structure of Petronius' work. It is an easy matter, in fact, to discover sources for various specific characteristics of the Satyricon; and it is likewise easy to see that Petronius must be credited with a considerable amount of originality in the handling of his basic literary model, whatever that may have been. In attempting to decide what this model most probably was, we ought to look for a form of literature which bears the most fundamental similarity to the Satyricon, not in the details of subject-matter, or even plot, so much as in the main tendency of the story and its more radical type characteristics. The fact, for instance, that the Satyricon contains a long description of a banquet is surely less significant in the question of its origin than the fact that it is, by and large, a narrative of adventure. And, when we have chosen that form of literature which appears to make the nearest approach to the Satyricon, we shall want to define as far as possible the originality of Petronius, and to account for the gap existing between his work and its assumed literary ancestor. The narrower the gap, and the more readily it may be explained in terms of literary growth and practice, so much the more probable will be our choice of the original model or logical forerunner. The chief difficulty with many of the suggested lines of descent lies in the fact that they postulate such wide gaps between the Satyricon and its supposed antecedents, and such radical innovation or reconstruction on the part of Petronius, as can scarcely be paralleled in literary history. This, I think, should be avoided if at all possible. We should keep to the historical and evolutionary method, even if it becomes necessary to assume a missing link.

For the sake of a little orientation, let me review briefly a few of the more important theories heretofore advocated.

The view that the Satyricon is a Menippean satire expanded into a romance (whatever that means) has been held by such a formidable array of scholars as Rohde,⁵ Ribbeck,⁶ W. Schmid,⁷ Hirzel,⁸ and more recently by J. Geffcken.⁹ None of these men, however, has had much to say on the subject beyond a few obiter dicta. They note the title of the work, Satyricon or Satirae, the mixture of prose and poetry as in Varro, and the recurrence of numerous themes common to satire. But it is hard to see how any of these similarities can be regarded as fundamental.

The title as given in the manuscripts ranges from Satirarum libri and Saturicon, or its intended equivalent Satyrici libri, through various obvious corruptions of these words. The best manuscripts have Saturicon;;¹⁰ and it is surely much more probable that an original Greek genitive Satyricon has been misunderstood and corrupted into the familiar Latin Satirae than, conversely, that Satirae has evolved into the less familiar but correct form Satyricon (sc. libri). The propriety of such a title is not hard to understand; it falls in line with the usual title of a Greek romance, such as (Suidas, s.v. lamblichus), (Heliodorus), etc.,¹¹ and probably meant simply a romance dealing with things of a satyrlike character. That is, in fact, exactly what we have in Petronius' work; for Priapus plays an important part in the plot, and the activities of Encolpius, Giton, and Ascyltus are pre-eminently satyrica in this sense, that is, phallic. The genitive ending in -on shows clearly enough that Petronius regarded his title as Greek, i.e., not derived from Latin satira; and this Greek word … regularly had the meaning satyr-like or pertaining to satyrs, whether applied to the drama or to anything else, from the earliest times to the latest.¹² The elder Pliny uses the word in this sense when he speaks of saturica signa, i.e., statues of Priapus (xix. 50) or saturicos motus (of certain birds, x. 138); and Plutarch applies it to men who resemble satyrs, either in outward appearance (Cato 7) or in conduct (Galba 16; Pericles 13). The title Satyricon (sc. libri) is therefore thoroughly appropriate to an obscene novel; whereas, Satirae, which regularly means a number of separate satirical essays, seems much less appropriate and more difficult to explain.

As for the recurrence in the Satyricon of numerous themes common to satire (a very loosely defined type), it should be remembered that many of these are also common to other forms of literature, and that even in their aggregate they are probably not much more numerous than the themes belonging to the mime,¹³ for instance, or to the Milesian tale (i.e., realistic novella). As Professor Mendell observes, it is natural that satire should have influenced the novel of Petronius as it did other forms of Roman literature, epigram, lyric, mime, history, and even epic (see Petronius' Bellum Civile); but the fact of this influence, which is also reflected in the prose-poetic form, does not go far toward convincing us that the Satyricon is essentially or primarily a satire, or that it owes its origin to satire. In the main narrative, which is what we have to explain, Petronius differs from Varro and the satirists in that he shows no evidence of moral seriousness. Everything is presented from a purely objective point of view, to all appearances merely for fun, and without any traces of the author's approval or disapproval. If Petronius had any ulterior philosophical purpose in describing the burlesque adventures of Encolpius, then he has concealed that purpose very effectively; for the tone is nearly everywhere gay and always unmoral.¹⁴ On the other hand, the brilliant harangues on literature and art, and the clever, often beautiful, poems that have been inserted into the main story undoubtedly represent the author's own serious thought and his best artistic effort. That these are formally subordinated to the burlesque narrative, and that they are put into the mouths of rascals, or even ridiculed at times, may be explained as due to Petronius' dislike of posing as serious or didactic. To do so, even in the guarded manner of Horace, would not only be distasteful to him as a sophisticated courtier, but, in the realm of poetry at least, even dangerous. Petronius must have known Nero well enough to beware of his jealousy. If he was to give expression to that poetic genius which he possessed in a greater degree than anyone of his age, he must not, like the ill-fated Lucan, profess to be a poet, but only a trifler. Accordingly, the Satyricon consists mainly in a purely burlesque and unmoral novel (a form apparently despised by ancient critics), while the artistic expression of the author is made to appear incidental and playful. Which of the two elements took precedence in the author's mind and was responsible for the writing of the book, we have, of course, no means of determining; but the composition of a long, burlesque novel, though it served a definite purpose and gave the author many an opportunity for self-expression, was probably no mere means to an end but likewise an end in itself. Such a performance rings true to the character of the cynical arbiter elegantiae as described by Tacitus—the man who regarded nothing more worth while than idle amusement and who, at the hour of death, "listened to no discourses on the immortality of the soul or teachings of the philosophers, but only to trivial songs and light verses."

Since the story part of the Satyricon has every appearance of being written primarily to amuse, we may conclude that it is not a satire, expanded and incidentally taking on the form of a romance, but rather a romance which has been influenced to some extent by satire. It is possible, of course, that this romance was the first of its kind, and that it was created on the basis of no better prototype than a Menippean satire; but the transition here seems too abrupt, and there are other forms of ancient literature which make a nearer approach.

The attempt to establish some sort of connection between the work of Petronius and the Greek erotic romances, though always inviting, was long delayed by the prevailing belief that the latter species did not come into being before the second century A.D. But this date for the origin of the Greek romance is now known to be wrong. The discovery of the Ninus romance on a papyrus which had become waste paper in 101 A.D. has shown very clearly that the erotic romance, as a type, must have been in existence at least by the middle of the first century A.D., probably much earlier."¹⁵ It is possible, therefore, that Petronius took his pattern for the Satyricon from the Greek romances. This hypothesis would explain in some measure the provenience of the general type, a story of adventure featuring two lovers whose experiences sometimes bear a close outward resemblance to those of the Greek hero and heroine. But, even so, we have still to account for the vast difference in nature between the two species of romance: the Greek is idealized and serious, while Petronius is realistic and burlesque.

Richard Heinze has attempted to explain this difference on the theory that the Satyricon was written as a deliberate parody upon the Greek romance.¹⁶ Most critics will admit that the Satyricon does contain parody on romantic love, but that this parody was the dominating motif and raison d 'etre of the entire romance is by no means clear. The parody is too poorly sustained. It appears to be merely incidental. In such a work as Heinze assumes the Satyricon to be, the long episode of Trimalchio's dinner would be quite out of place. Then, too, we find parody on the epic as well as on the romance. Encolpius more than once compares himself to Ulysses and, like the Homeric hero, he, too, is pursued over land and sea by the wrath of a deity, in this case Priapus. But the authors of genuine parodies, such as Lucian's True History, or the Battle of the Frogs and Mice, never leave us in doubt about the thing parodied. From beginning to end, the tendency is everywhere obvious and the parallelism in technique and motivation carefully sustained. Of course it is difficult to say just what constitutes parody. It is easy to agree with Heinze in general when he says that the Satyricon stands in about the same relation to the serious Greek romance that comedy or satyr-drama does to tragedy. In both cases we observe a sort of reaction. But it is hard to believe that the parallelism between the two romances was so close as between the two kinds of drama, or that in either case the origin of the comic type is to be explained as due to intentional parody. The Satyricon is to the Greek romance no more than what Gil Blas or Lazarillo de Tormes are to the romance of chivalry.

The most recent writer on this subject is Professor C. W. Mendell (loc. cit.). Mendell rejects the theory of a deliberate parody and, regarding the machinery of the plot in the Satyricon as its most essential element, maintains that Petronius' work represents merely an advanced stage in the development of the serious erotic novel. He thus fails to account for the burlesque and unmoral tendency except in so far as it may be due to the influence of satire and the sophisticated age and surroundings in which the Roman writer lived. To this it may be objected that, besides the inherent improbability of any literary type radically changing its primary tone and tendency except by parody, there is no evidence that such a transition took place, unless it was due entirely to the wilful invention of Petronius. The latest of the Greek erotic romances written long after Petronius are quite as idealized and as serious as the earliest. They show no tendency whatever to replace heroes with anti-heroes, as Petronius has done, nor to change the serious tone to the comic. As ancient comedy appears to have sprung from a different type of origin from that of tragedy, and as both types remained distinct throughout antiquity and characterized by a different tone and tendency, so, I believe, the comic unmoral novel, though formally influenced by the conventions of the serious romance, must have been comic or at least realistic at the start.

The history of the novel in later times would suggest this. The humorous, realistic, and somewhat unmoral history of the roguish Gil Blas, though less exaggerated, is fundamentally the same kind of story as the Satyricon. Yet its origins are not to be sought in the serious romance of chivalry but directly in the rogue stories of Spain, such as that of Lazarillo de Tormes. And these rogue stories themselves, if they did not, as seems most probable,¹⁷ result from the grouping about one character of numerous stock tricks and facetiae, at least did not grow out of the serious romances.

But to return to Petronius. The question as to whether or not the Satyricon was preceded by other romances of a burlesque or realistic or picaresque nature has generally taken the form of a dispute as to whether the lost Milesiaka of one Aristides, translated by Sisenna, was a collection of stories or a continuous romance like that of Petronius.¹⁸ The question is at least an open one; but since the ancient testimonia are quite ambiguous on this point, we shall do better, I think, to leave Aristides out of the reckoning altogether and to admit that, so far as explicit testimony goes, we cannot be absolutely sure of the existence of any comic romance prior to the time of Petronius. But lack of testimonia in the field of ancient fiction means very little; and it would not be at all surprising if many comic romances, of which we now have no knowledge, were in circulation in the days of Petronius.¹⁹ When I say "comic romance" I do not mean a romance resembling the Satyricon in all its wealth of realistic tableaux and numerous side shows, but rather a straightforward story of manifold adventure related chiefly for the sake of the fun and in the spirit of burlesque. That such a romance existed in Greek before the time of Petronius appears to me to be extremely probable, and it is only when we make this assumption that the origin and peculiarities of the Satyricon can be explained satisfactorily and in accord with the facts and tendencies of ancient literary history.

We know that a comic romance did actually evolve in Greek literature. The (Loukios e Onos] of Lucian, familiarly known through Apuleius' interpolated version as the story of the Golden Ass, is just such a romance as I have in mind. The foundation on which it was built was a short folk-tale which related the transformation of a young man into an ass as a punishment for some folly or for some offense against the witches.²⁰ The author of the [Onos], or rather of its original,²¹ has taken this simple situation and made it into a "romance" by giving it an introduction and a conclusion and by adding to the number of the experiences which the young man undergoes. The added experiences in this case were suggested partly by Aesopic fables and proverbs relating to the ass; but there is also, as in Petronius, obvious borrowing from the epic, the comedy, the erotic romance, and the mime.²² Now the basic story of the Satyricon is closely analogous, both in formation and tendency to the [Onos]. It consists mainly in a series of comic episodes suggested by or taken bodily from various convenient sources, especially the mime, and related smoothly and, no doubt, as in the [Onos], with much originality and invention, as the experiences of one man.²³ Without reaching a climax such a story might be prolonged indefinitely. The only logical end would be the death of the protagonist or want of comic experiences to assign to him. The adventures of Encolpius were probably represented as the result of his having offended Priapus, while those of Lucius resulted from his own fatal curiosity about magic. In both cases the cause of the adventures serves as a loose framework on which the episodes are hung, and which gives them the only thread of unity they possess, apart from the biographical form. Other romances of the same type may have had a different framework. This framework merely supplies the want of a plot; and the fact that it differs in the two romances is no more significant than the fact that Plautus' Menaechmi turns on a different plot from that of his Trinummus.

Nor do I believe that the erotic element is any more essential. The pairing of Encolpius and Giton as lovers with Ascyltus as a foil appears to be merely a device for creating comic situations. Their mock love affair involves no sustained dramatic suspense as in the erotic romances properly so called, nor is the interest here, as there, at all psychological. Unlike the picturesque characters of serious romance, Encolpius and Lucius, as human souls, attract neither our interest nor our sympathy. They are far from being even clever rogues. They are merely the vehicles of burlesque. The primary and ever present purpose of both the Satyricon and the [Onos] is simply to amuse the reader by the objective presentation of consecutive comic scenes. And herein the type is defined. So long as the episodes are presented for the sake of the fun, and not for love's sake or for the sake of realism, they may be either erotic or non-erotic, real or imaginary. Erotic scenes lend themselves very readily to burlesque, and it is for that reason no doubt that they figure so prominently; but their value is purely comic, and their importance no greater than that of numerous other burlesque scenes of a different nature.

Likewise, the presence or absence of realism in the primary situation seems unimportant. The difference between the [Onos] and the Satyricon in this respect is merely a difference in the particular devices employed toward the same end, namely, to create a potentially comic situation or framework. Such matters of plot-technique, or choice of theme, are variable within the limits of universally recognized and well-defined literary genres. Aristophanes' Peace and his Acharnians are both comedies, and written for the same political purpose, though the former deals with a frankly impossible situation, and the latter with a situation within the range of possibility; so, too, with the Amphitryo of Plautus as compared with the other plays; and so with Horace's Satires. Without differing in what we call type, they may deal with either possibilities or impossibilities. The supernatural incidents in the [Onos] are not told as such for their own sake, but in a spirit of burlesque, and they are strictly subordinate to the author's main purpose—comedy and amusement. As soon as Lucius becomes an ass his experiences are as realistic as those of Encolpius. Burlesque naturally adheres to the commonplace; but any device may be employed to support it.

Apart from the formation and purpose of the two stories of adventure, we note many other points of similarity. One of the most peculiar features of the [Onos], as of the Satyricon, consists in the absence of any moral personality in the leading characters. Outside the popular novella with which the [Onos] is closely associated in origin, one looks in vain for this strange quality in almost any other form of ancient literature. The conventional rogue, or even the parasite, usually acts with some spiritual energy, and has some kind of self-respect; but the peculiar thing about Lucius and Encolpius is that they relate the most extravagant and ironical farces all at their own expense. Provided it be comic, there is no act or predicament, however absurd or humiliating even to a rogue, to which they will not readily confess. The things that Lucius tells about himself, though humiliating, are generally less debasing than the experiences of the Roman Encolpius; but the tendency and the odd effect are exactly the same. In most narrative literature the episodes are adapted in some measure to suit the character, typical or individual, of the protagonists; but in the [Onos] and in the Satyricon the episodes exist for their own sakes, and the persons who are made to enact them tend thereby to become mere puppets. Accordingly, Lucius and Encolpius are fictitious persons representative of contemporary society, as is usual in comedy; whereas, in the erotic romance the characters generally belong to history or local legend and their experiences, except in the latest of the romances, that of Achilles Tatius, are represented as taking place in a far-off past. It is probably mere coincidence that both Lucius and Encolpius are educated Romans, the former an author, the latter a rhetorician; but in the erotic novel, the characters are never either Roman or literary. The [Onos] and the Satyricon are also alike in that they are both independent of the conventional geographical background of the erotic novels, for the latter rarely or never take us so far west as the home of Trimalchio, nor so far north on the mainland of Greece as the setting of the [Onos]. Like the Satyricon, the [Onos] has the autobiographical form, contains many comic reminiscences of the epic and the erotic romance, and also many similarities in the motives and situations. To enumerate these here would take too long, but I think I have made it clear that the basic story of the Satyricon is identical in type essentials with the story of Lucius.²⁴

The question now arises, Was this type originated by Petronius and imitated, as scarcely any other Roman type was, by the Greeks, or did both representatives spring up independently? Neither of these suppositions seems at all plausible. In nearly every department of literature the Romans were inspired by Greek models Why not here too? Is it not more likely that the first specimen of a comic romance was simple and straightforward like the [Onos], rather than a story like that of Petronius, which is imbedded, and almost lost, in a maze of digressions and embellishments of every sort? Petronius' work in sixteen books or more has every appearance of being a developed rather than a primitive (i.e., the first) specimen of the comic romance; and it is therefore very probable that the Satyricon was preceded by other comic romances of the more simple type, presumably in Greek.²⁵

This sort of novel may have originated in the same way as the [Onos], that is, from the expansion of popular tales that were either ironical or humorous at the start or potentially so by virtue of their subject-matter. A fairy tale of real intrinsic beauty is likely to retain its ideal character perennially; but a realistic tale, or a naively superstitious one, however serious it may be at the start, owing to its essentially homely character, is apt to become ironical or burlesque as society becomes more sophisticated. One may see this process at work in Apuleius' tale of Socrates and Aristomenes (Met. i. 5-20). It is true that not many of these popular tales were likely to become expanded into "romances," because the single incident with which they usually dealt often offorded no good framework for further episodes. But when fancy has once changed a man into an ass, nothing is more natural than that his recorded experiences in that form should be augmented, not indeed by popular repetition—for in a novella of this sort where the protagonist is variable and of no personal importance or identity there can be no popular biographical interest—but by the conscious literary effort of a writer already familiar with the Odyssey and with the serious prose romances. Besides the ass-story we may imagine others equally capable of expansion. The clever thief who stole from the treasure-house of Rhamsinitus (Hdt. ii. 121) is waging a war of wits with the king until such time as the latter pardons him. The situation here provides a natural framework into which any number of episodes might be worked without any organic readjustment. The difference between a story of this kind and the [Onos] is purely quantitative and artistic. We call the former a novella instead of a romance because it is popular rather than literary and because it contains fewer episodes; but the two types are structurally identical and spiritually very closely akin. Since the gap between them is very slight, it may well have been bridged by writers who lived before the time of Petronius. Such a development is more to be expected in an age that witnessed the elaboration of the mime, the new comedy, and the serious bourgeoise romance, and in which the realistic novella emerged from oral tradition into literature, than in the later and much less creative period of the second century A.D. (date of the [Onos]). After a few novels had been written on the basis of folklore plots, it would be very easy for any writer to create a plot of his own on the same simple principles, and this is probably what Petronius has done, possibly others before him.…

But there is more to account for in Petronius than the basic story. What shall we say of the poetry, the declamations, the inserted fabulae, and the vivid and detailed representations of scenes from real life? I believe most of this is to be strictly Roman and to have originated with Petronius. The practice of dressing up Greek models with various embellishments and digressions and of departing more or less from their conventional standards in form and range of content was apparently not uncommon among Roman writers. They wanted to contribute something original. Their attitude is somewhat typified by that of Phaedrus as expressed in the Prologue of his second book (1. 8): Equidem omni cura morem servabo senis (Aesopi)l sed si libuerit aliquid interponere, | dictorum sensus ut delectet varietas, | bonas in partes, lector, accipias velim; and at the beginning of the third book (1. 38): ego porro illius semita feci viam | et cogitavi plura quam reliquerat; and his fifth book professes to be entirely original. Likewise, Ovid's Metamorphoses, though based in large part on Greek collections of myths, is undoubtedly original in respect to its Rahmenerzdhlung, which cannot be paralleled in Greek, and to its mixture of epic, lyric, historical, and philosophical passages. These combinations and the arts by which the various myths are formally strung together may be accredited to the invention of the Roman artist himself. If we bear in mind the great amount of heterogeneous matter that Apuleius in his Metamorphoses has crowded into the framework of the straightforward Greek story of Lucius, we shall have no need to wonder at the disc'ursiveness of Petronius nor at his apparent aberrations from the norm of his Greek predecessors. The two Roman writers have treated their Greek models in almost exactly the same manner. Let us consider the various features separately.

Petronius differs from the "[Onos] and from Greek writers generally in the degree of realism he employs. By "realism" I mean the concrete, detailed, and vivid representation of scenes from ordinary or low life, often described for their own artistic value. We meet with some of this realism in the Greek comedy, in Greek mimes, and elsewhere; but on the whole it is less vivid and more incidental. Greek writers are likely to make less of it than the Romans.²⁶ At any rate, this is the case in the story of Lucius; for the Latin version of Apuleius contains many a graphic sketch of persons and things which, in the Greek version, were alluded to in more general terms or briefly dismissed.²⁷ It is probable, therefore, that the extreme realism of Petronius is Roman; and that just as the Romans never, so far as we know, developed an idealized romance, so probably the Greeks never had a truly realistic one, that is, realistic in the same degree as Petronius. The realism in the "[Onos], which I assume to be more or less typical of that of the lost novels of the same type, is purely incidental to the narrative; whereas in Apuleius and in Petronius it is often paraded for its independent value; cf. Met. ix. 12-13; and the description of Trimalchio's dinner.

That which has been said of the difference in point of realism between the Roman and Greek novels applies also to character-drawing; for the characters in Apuleius, as well as in Petronius, are described far more vividly and realistically than they are in the "[Onos] or in any other Greek romance. Since Apuleius has made a great advance in this respect over his Greek original, it is easy to believe that Petronius has done likewise.

Along with concreteness we observe greater obscenity, and more of it in Petronius than in the "[Onos], or, in fact, in any known Greek work. The Greek tends to keep it subordinate, the Roman to give it greater prominence, and often to make it coarser by representing it more concretely. This again could be amply illustrated from Apuleius, to say nothing of Catullus and Martial in contrast with the Greek Anthology (quantitatively and qualitatively).

Another feature worth mentioning is the insertion of short stories more or less independent of the plot. Such stories are those about the werwolf, the matron of Ephesus, and that of Eumolpus in chapters 85-87. Apuleius has inserted seventeen or more such independent stories into his translation of the Greek Luciad, and Ovid tells us that Sisenna added ioci to his translation of Aristides. These ioci, as Professor Mendell observes, may have been short stories; but in any case it is important to note that Sisenna did add some embellishments of his own, and that it is only in the Roman novels, never in the Greek, that independent stories are inserted.

The long poems on the fall of Troy and on the Civil War, respectively, and the lengthy discussions of literature and art, interrupt the progress of the main narrative and are introduced chiefly for their intrinsic interest. At first thought, it would seem difficult to parallel this sort of procedure elsewhere, but here again a comparison with Apuleius proves instructive. Apuleius does not indeed introduce long poems, doubtless because he was less interested in poetry than Petronius; but he does introduce philosophical digressions and ornate descriptions, and these, like the digressions of Petronius, stand apart as artistic units, treated for their own sake and retarding the narrative. The difference in nature or length between the separate artistic entertainments of Petronius and those of Apuleius are due only to individual taste. Petronius as arbiter elegantiae at Nero's court was interested in literary criticism and matters of taste. It is not surprising, therefore, that he felt the challenge of Lucan's Pharsalia and was tempted to imitate or rival or parody that poem in his Civil War. Apuleius, on the other hand, being more interested in his prose style, exercises his talent in describing the house of Byrrhena, or the robber's cave, or in writing an essay on the beauty of human hair, or describing in ornate prose the ceremonies in the worship of Isis and Osiris, all of which he has added of his own accord to the Greek original, and which are unessential to the story. Likewise, the shorter poems in the Satyricon rarely advance the action, but, like the others, seem to be introduced for their independent interest. Sometimes, of course, they illustrate a point in the text, or a situation, quite effectively and humorously, but they seem to be added for the sake of embellishment rather than as a matter of form. Many of them are of surprising beauty and stand in the same odd contrast to their coarse surroundings as do the artistic effusions of Apuleius.

When Apuleius puts the beautiful story of Cupid and Psyche into the mouth of a villainous old hag, he is doing the same sort of thing that Petronius does when he puts elegant criticism into the mouth of Eumolpus. As a mystic, Apuleius is induced to leave out the original burlesque ending of the ass-story, and to substitute therefor a twenty-five-page chapter describing in a solemn, religious tone the majesty of Isis and Osiris. As a platonic philosopher and a Carthaginian senator, he puts into the mouth of Lucius a page of denunciation against corrupt judges and an eloquent eulogy of Socrates, but at the end he remarks: "Behold, shall we suffer an ass to philosophize? I return whence I left off, to the main story" (Met. x. 33). In the same way, Petronius has allowed his degenerates to philosophize on literature and art, or in fact on any of the numerous subjects in which he is interested, and which may add to the reader's entertainment.

These embellishments probably do not belong to the comic, unmoral novel as a Greek type but are added in accordance with the fancy of the individual Roman writer; and this in turn will be qualified by the age in which the author lives and its cultural and literary background. Much of the subject-matter of Petronius, as well as the discursive nature of his writing and his love of variety, shows the influence of Roman satirè; but it is not as a moralist that the arbiter elegantiae is interested in society but as an entertainer.

To sum up, I believe that the basic story of the Satyricon was patterned after, or at any rate preceded by, some straightforward comic narrative like the "[Onos]; and that the criticism of art and literature, the poetry, the character-drawing, and the realistic tableaux are due to the originality of Petronius.

Notes

¹ Read at the meeting of the Ohio Classical Conference at Delaware, October, 1923, under the title, "Petronius and his Greek Sources."

² Cf. chap. 139:

³ Chap. 19, omnia mimico risu exsonuerant. I agree with Preston (Class. Phil., X, 261) and Thomas (Petrone, p. 213) that the keynote of the Satyricon is struck in this passage. Preston observes very justly that Petronius "conceived of himself primarily as a [gelotopoios]" and that "nowhere else in Latin literature is such a premium put on laughter."

⁴ These are discussed in an interesting article by Professor F. F. Abbott, Class. Phil., VI, 257 ff.

⁵Der Griechische Roman,³ p. 267.

⁶Röm. Dichtung², III, 150.

⁷Neue Jahrb. f d. kl. Alt., XIII (1904), 476.

⁸Der Dialog, II, 37.

⁹Neue Jahrb. f d. kl. Alt., XXVII (1911), 485.

¹⁰ That is, BDEFG_pt. See Bücheler's Preface in the edition of 1862, pp. xiv, xxv-xxvi, and 2. Petronii arbitri satirarum 1. is the reading of P; but P seems to be the only MS in which any form of the noun satira occurs.

¹¹ Cf. C. W. Mendell in Class. Phil., XII (1917), 168; and Suidas.… These titles are cited in the nominative, of course, but they probably appeared on the title-page in the genitive followed by [logoi] or [biblia] plus a numeral, or with the numeral alone. Strictly, the work of Petronius should be cited not as Satyricon, but as Satyrica.

¹² The meaning "satirical," or "belonging to satire," appears to be late and to have originated with grammarians who associated Roman satire with the Greek satyr-drama; so Lydus, De Mag. 41. After a prolonged search through the lexicons and indices verborum, I am unable to find any instance earlier than Lactantius or the scholia on Juvenal i. 168, in which the word satyricus has reference to satire, either in Greek or in Latin. That it may have been so used by grammarians in the time of Petronius is not improbable; but the other meaning was certainly common, and of much longer standing.

¹³ See the parallels listed by M. Rosenblüth, Beiträge zur Quellenkunde von Petrons Satiren, pp. 36-55.

¹⁴ Professor Abbott observes with a great deal of truth that much of what appears to be satirical in Petronius is so only because we are setting up in our own minds a comparison between the abuses described (perhaps merely for fun) and the requirements of good taste. The subject-matter of the Satyricon, like that of the realistic novella, by its very nature may be regarded as constituting a satire on society; but this does not mean that the author is a satirist, if by "satirist" we mean one whose chief purpose, like that of Varro, Horace, or Juvenal, is to criticize society from an ethical point of view. For this implies either an attempt to correct, or moral indignation or reaction to things as they are. But the only inference, if any, that Petronius by his tone would encourage us to draw seems to be that society is incorrigible and not worth worrying about; and that it is the part of wisdom not, like the satirists, to carp at conditions which are sadly inevitable, but to look only for amusement in the comedy of human life. Petronius is a cynic; but his cynicism is not that of the school, or of Menippus, who would scale heaven for philosophical truth. It is deeper and more somber; it springs evidently from a profound though latent pessimism, from the cosmic disillusionment of the man of the world.

¹⁵ See the scholarly monograph of B. Lavagnini, Le Origini del Romanzo Greco, Pisa, 1921. Lavagnini shows in a very convincing way that the romance evolved from the elaboration, popular and historiographical, of local legends in Hellenistic times; cf. AJP, XLIV, 371 ff., and, for the early date, Mendell, op. cit., pp. 161-62, 165; and W. Schmid in Rohde's Gr. Roman,³ p. 610.

¹⁶Hermes, XXXIV (1899), 494-519.

¹⁷ See Chandler, Romances of Roguery, pp. 6 ff.

¹⁸ For the views of the leading disputants on this subject see the summary of Rosenbliith, op. cit., pp. 87-90. The most important ancient testimony is to be found in the following passages: Ovid, Trist. ii. 413-14; ibid. 443-44; Ps. Lucian, Amores 1; Plutarch, Crassus 32. From these passages, and from a few very meager fragments, the most that can be inferred with certainty about the [Milesiaka] is that they were obscene and partly at least in prose. That they may have been partly in verse also has been inferred by Norden (Antike Kunstprosa, II, 756) from the words nocte vagatrix quoted by the grammarian Charisius from Sisenna's translation. The plural title may mean no more than it does in Lucian's [Alethe diegemata], or in the [Aithiopika] of Heliodorus.

¹⁹ The ancient literary critics evidently regarded this kind of writing as trivial and beneath their serious consideration. Hence they tend to ignore it. That it was popular, there can be no doubt (cf. Jerome in Biucheler's Petronius, p. 243); but it must have circulated rather among laymen than among men of literary profession. The novels of Petronius and Apuleius are mentioned occasionally by ancient writers, but generally in a tone of disparagement. Outside of Photius and Suidas, references to Greek romances are extremely rare and meager; and even in these encyclopedias you will look in vain for mention of Longus or Chariton, whose novels would be quite unheard of were it not for the survival of their manuscripts. Likewise, the two erotio novels mentioned by Suidas under the name of Xenophon (cf. supra) are apparently mentioned by no other ancient writer. Moreover, some of the ethnographical titles that have come down to us and are generally believed to refer to historical works, may in reality have been the titles of romances. Thus, if Suidas had not added that [estide ton panu aiskhron] we would assume that the [Rodiaka] of Philip of Amphipolis was history; but the descriptive remarks of our lexicographer, and a casual reference in Theodorus Priscian (Res Medicae, 11), make it clear that the book was erotic fiction; and who knows but that it antedated Petronius? Along with Aristides, Ovid (Trist. ii. 415 ff.) mentions two other naughty books whose authors were not exiled:

Nec qui descripsit corrumpi semina matrumEubius, impurae conditor historiaeNec qui composuit nuper Sybaritica fugil.

Either one of these books may have been a continuous romance like that of Petronius, and the term historia strongly suggests this; for the usage cf. Mendell, op. cit., pp. 163-64; Propertius, ii. 1, 13-16; see also Burger in Hermes, XXVII (1892), 354-55. That they were written in a humorous vein seems probable from the analogy of the stories of adultery in Apuleius, Petronius, and elsewhere.

Robert, Hermes, XXXVI (1901), 364 ff., believes that certain wall paintings found in the casa Farnesina at Rome represent scenes from a picaresque romance prior to Petronius. But in this, too, there is no certainty.

²⁰ K. Weinhold (Sitzungsb. d. konigl. Preuss. Acad. d. Wissen. [1892], pp. 475 ff.) points out eight analogues to this story in the folklore of Europe and India. He summarizes as follows: Das Urgeschichtchen mag so gelautet haben: einjunger Mann kommt mit Frauen in zu vertraute Beziehung, und wird zur Busse in einen Esel verwandelt, dem gewisse seiner Anlagen entsprechen. Nur sein Ausseres, nicht seine innere Natur wird von der Verwandlung betroffen. Er hat ein muihsames Leben zu fuhren, bis es ihm gelingt, die Krauter zu geniessen, welche bestimmt sind, ihn zu entzaubern. For other legends, less typical, to be sure, than those discussed by Weinhold, but dealing with the same theme, see the story of Peter the Huntsman, in Grimm's Fairy Tales and that of the rogue Ali of Cairo in the Arabian Nights (Burton's translation, VII, 197-99). Cf. also my dissertation, pp. 43 ff.

²¹ The [Onos] is an epitome of a longer work of the same nature entitled [Metamorphoseis] and ascribed by Photius (Bibl. 129) to one Lucius of Patrae (the ass in the story). The real author of this lost original, however, was probably Lucian; cf. Perry, op. cit., pp. 59 ff.

²² For Aesopic motifs in the [Onos] see Crusius in Philologus, XLVII (1888), 448. Epic parody and reminiscences are listed, though not exhaustively, by Neukamm, De Luciano Asini Auctore (Leipzig, 1914), pp. 92-93, who also points out the influence of comedy (pp. 94, 87-88). Several matters in the [Onos], such as the adventure with the robbers, their cave, their plans for torturing (chap. 25; cf. Xenophon of Ephesus iv. 6), the [daimon baskanos] (19), the resolve of Lucius to commit suicide rather than become a eunuch (33), the setting up of [anathemata] at the end (cf. Longus and Xenophon, ad. fin.), as well as occasional stylistic features (e.g., the soliloquies of Lucius in 5, 15, and 23; cf. Chariton vi. 6; Xenophon ii. 10, iii. 5; Heliodorus ii. 4; Achilles Tatius iii. 10) remind us of the erotic romance. For the mimic motifs in the [Onos] compare chap. 51 with the statement of Suetonius in Nero 12 and Martial in Liber Spect. 5. See also Rosenbliith, op. cit., p. 65 (top). In commenting on the boisterous theatrical element in Petronius, Preston (loc. cit.) observes that the humor of an incident is seldom left to make its own appeal to the reader, but "we are told that it provoked 'gales of laughter' or 'bursts' of applause." …

²³ By this I do not mean to imply that there was ever any progress of development from collections of separate stories or Schwtnke to novels like that of Petronius. This is the view of K. Burger (Studien zur Geschichte des' griechischen Romans [Erster Teil, 1902], pp. 20 ff.) and of Schissel von Fleschenberg (Entwicklungsgeschichte des griechischen Romans [Halle, 1913], pp. 3 ff.), who assume that Aristides' Milesiaka was some kind of a collection—Rahmenerzählung, according to Schissel—and that there was a tendency among writers of such collections to weld together the separate stories more closely so as to give them some inner unity, and that finally they came to be told smoothly as the experiences of one man. To me, this theory of development seems very improbable; cf. the remarks of W. Schmid in Rohde's Gr. Roman³, pp. 607 f.

²⁴ This fact is recognized, though somewhat vaguely, by Bürger (op. cit.) and by Collignon (Étude sur Pétrone, p. 49), neither of whom has made an adequate and discriminating comparison.

²⁵ Cf. the remarks of Leo, Die Kultur der Gegenwart, Teil I, Abt. 8 (Leipzig, 1912), p. 459: Die Form des Schelmenromans, die das Buch hatte, war gewiss in den Unterschichten der griechischen Litteratur vorhanden; was solche Produktion wert ist das hängt ganz von der Persönlichkeit ab, die das Ihrige in die Form hineinlegt. Wilamowitz (op. cit., p. 190) is of the same opinion.

²⁶ Cf. the interesting remarks of F. A. Wright, AJP, XLII, 169: "The habit which the Roman poets have of working up a long passage from a few lines in some Greek original by the addition of a mass of realistic details deserves more study than it has yet received." He cites several examples from the works of Vergil, Ovid, and Horace.

²⁷ A fuller discussion of the various phases of Apuleius' originality in the Metamorphoses will be found in an article of mine which appears in TAPA, LIV. 196 ff.