The five extant Greek novels appear to be a uniform set. In these so-called “ideal” novels, a young Greek couple meet, fall in love, perhaps marry, but then are separated for the duration of the text, and have dangerous adventures in foreign lands before they are reunited, presumably to live happily ever after. The three Roman novels display greater narrative variety; there are no young lovers (unless we count the homosexual pair in the Satyricon), but magic, chicanery, and subversive behaviors, to the extent that these texts are sometimes positioned as inversions or parodies of the more serious Greek set (Petronius’ novel, if Neronian in date, would be earlier than any Greek novel we know). Theories of the ancient novel depend, in the first instance, on how the field is bounded; are these eight the sole representatives, or should a wider array of fiction, the Alexander Romance or the Ps-Lucianic ass-tale, for example, also be included? However, we have incipient circularity; what you include depends on some a priori (if subconscious) notion of what the novel was, but the included texts in turn dictate how you formulate an idea of the ancient novel. If the five ideal Greek novels were the only novels we had, then the question might be irrelevant but, in fact, fragments or epitomes of prose narratives survive from antiquity that do resemble the extant novels in so many particulars that they have come to constitute a subset of the field; the “fragmentary novels” in turn generate their own set of categorizing questions. For collections and discussions of fragmentary novels, see Kussl 1991; López Martínez 1994; Morgan 1998; Stramaglia 2000; and Stephens and Winkler 1995.
Of the fragments we (Stephens and Winkler 1995) printed in Ancient Greek Novels: the Fragments, there were sufficient textual remains of Metiochus and Parthenope, Ninus, Antonius Diogenes’ Unbelievable Things beyond Thule, Iamblichus’ Babyloniaka, and Lollianus’ Phoinikika to ensure we were working with sustained narratives; these sufficiently resemble the extant Greek and Roman novels to allow for them to be assigned to the same presumptive category. Several smaller fragments have scenes that suggest they too belong within this group, though their brevity deprives them of any narrative complexity that they might have had and makes the call riskier: these are Herpyllis, Kalligone, Antheia, Chione, Sesonchosis, and Iolaus. At least 15 other fragments may well belong to novels or novel-like narratives, but they are far too small to attribute with certainty. In what follows, I will concentrate on the five large fragmentary texts and the six smaller ones by sketching out what we know of their plot lines and how they intersect with the surviving novels and other contemporary writings. However, first it is worth looking at the patterns of survival for this material as a group.
We find two venues: Photius, the ninth-century bishop of Constantinople, provides plot summaries for Iamblichus and Antonius Diogenes along with summaries of Heliodorus and Achilles Tatius, so we can be sure that the good bishop at least imagined that these four were all similar types of narrative. What we can also infer from comparison with Photius’ retelling of the Aithiopika is that he preferred linear narrative to the tergiversations of the novel’s actual plot. Therefore, the summaries give us the general idea, but not necessarily the events of the plot in the order in which they were written. These two are also represented in ancient sources: the Suda has numerous quotations from Iamblichus, while Porphyry and John Lydus both excerpt from Antonius Diogenes, a circumstance which suggests that these fictions had some stature for later Greek scholars. The second source for fragments is from Roman Egypt, where the aridity of the soil has yielded (to date) over 10,000 remnants of Greek literature. The majority of them belong to the Roman period, and range in date from the first through the fourth centuries CE, in fact, to the time when the ancient novel first appeared upon the scene. Today, the number of fragments of potential novels is still fewer than 50, and the conclusion that such texts (at least in Egypt) belonged to elite readers is difficult to escape (Bowie 2003; Stephens 1994). Their formatting and writing style is indistinguishable from papyri of Plato or Thucydides. In contrast, the number of fragments of Homeric texts (in the thousands) or Christian texts (well into the hundreds) provides a better insight into what might have constituted “popular” genres in antiquity. Although the small samples make inferences risky, it does appear that some novels were more popular than others—there are six fragments of Achilles Tatius, four of Chariton, at least four of Antonius Diogenes, four of Metiochus and Parthenope, and three each of Lollianus and Ninus. In contrast, Longus and Xenophon of Ephesus are not represented at all, and there is only one fragment of Heliodorus (which was originally identified as “Egyptian history”). This matches the distribution for other later Greek writers like Plutarch (9), Aelius Aristides (7), Libanius (3), or Lucian (2).1
The usual chronology for the extant Greek novels locates Chariton as the earliest and the Aithiopika as the latest, written sometime in the third or early fourth century CE. The fragments, which are dated on the basis of internal references, linguistic features, and handwriting, show a similar range (see Stephens and Winkler 1995 for additional details for each fragment; for fragments published later, see Hägg and Utas 2003 or the relevant volume of the Oxyrhynchus Papyri). Metiochus and Parthenope and Ninus are to be located within the first centuries BCE–CE. Iamblichus and Lollianus are probably from the second half of the second century CE, whereas the Sesonchosis fragments seem to be quite late (third–fourth centuries CE). A recent discussion by Ewen Bowie (2002, 58–60) makes a very strong case for thinking that Antonius Diogenes should be dated to the second century, if not before. One feature common to most of these fragmentary texts is that the plots of them all seem to diverge from the canonical five, either in the type of their main characters, or their geographic range, or their affinities with non-Greek narrative fiction.
The most interesting in this last respect are Metiochus and Parthenope, Ninus, Sesonchosis, and Iamblichus. Metiochus and Parthenope has the most remarkable textual career of any ancient novel. From our four Greek fragments, we can glimpse the basic outline of the plot. Like Chariton, it seems to have been quite early in date: the language and style suggest a date no later than the first century CE (the characters of Metiochos and Parthenope were known to Lucian, writing in the mid-second century). Moreover, its fictional antecedents are to be found in Greek history: the protagonists are respectively the son of the Athenian general, Miltiades (Hdt 6.39–6.41), and the daughter of Polycrates of Samos (Hdt. 3.124). Metiochus comes to Polycrates’ court in refuge from his stepmother, where he meets Parthenope. Their courtship does not quite follow the lines we find in later novels, though. Our protagonists engage in philosophical discussions rather more than lovesick gazes. At some point, Parthenope leaves home (possibly sold into slavery?), and experiences a series of adventures in Sicily and South Italy. Metiochus also wanders, probably in search of her, but whether they are ever reunited cannot be determined from the fragments that survive. However, this Greek fiction is not the end of their story: Thomas Hägg and Bo Utas (2003) have identified a Persian novel, Vamiq u ‘Adhra (“The Lover and the Virgin”), based on the Greek, and there also seems to have been a Coptic martyrdom (of St. Bartanuba) descended from the original. Thus, Metiochos and Parthenope are implicated in a series of exchanges, transcultural and inter-faith, that provides us with a glimpse of the potential and complex interactions in which fiction can operate. One divergence from the norm of the canonical five is that none of these later versions suggest a happy ending for the Liebespaar, and Hägg and Utas (2003, 250) even conjecture that Parthenope may have met her death before or instead of marriage with Metiochus. Whatever their fates, the story seems to have been extremely popular. Two floor mosaics have been discovered that name both figures, one in Daphne in Roman Syria from the late second or early third century, and a second from Zeugma-Belkis on the upper Euphrates. Lucian in On Dancing indicates that it was a subject for a mime (2.54), so the mosaics need not necessarily represent the novel, but a collateral version of the story. Still, the novel, the mime, the mosaics, the martyrdom, and finally the appearance of the novel in a Persian form, all attest to a wildly popular narrative that (unlike the canonical five) made its way quite successfully into the Persian, Arabic, and Coptic imaginations.
Sesonchosis provides us with insights into a different type of cultural exchange. In addition to two Greek fragments, several demotic fragments indicate again a trans-cultural exchange, though in this case the exchange may have been limited to Egypt, and perhaps reflect a Greco-Egyptian milieu for its composition. The Greek fragments are late (third–fourth centuries CE), and the novel’s origins, if it is in fact a novel and not a prose rendition of an Egyptian tale, seems to derive from demotic Egyptian accounts of a native pharaoh—Sesonchosis, who is called Sesostris in Herodotus (Widmer 2002). This figure and his exploits as a precursor of Alexander are well attested in Greek sources: besides Herodotus, he appears in Apollonius of Rhodes’ Argonautica, Diodorus Siculus (in material probably taken from Hecataeus of Abdera), and is prominent in the Alexander Romance. The pieces we have come from a scene between father and son (P.Oxy. 15. 1826), a description of a battle against the Arabs (P.Oxy. 26. 2466), and an encounter between Sesonchosis in disguise and a young woman to whom he has apparently been betrothed, although they have never met (P.Oxy. 47. 3319, part of same roll as 26. 2466). The battle scene is one of Egyptians routing Arabs, probably on the eastern edge of the Delta:
The Arabs, seeing themselves growing fewer each day and the Egyptians gathering still greater numbers as men came in from the other nomes, were so (effectively) routed that not the fifth part of their army saved itself, but some being pursued and others fleeing were trampled by one another (συμπα[τ]ε[ι̑]σθαι). (P.Oxy. 26. 2466. 7–16)
The demotic fragments tell of the military campaign of pharaoh Amenemhet and his son Sesostris against the Arabs, and these narratives, in a general way, may have been known to Diodorus’ sources, Hecataeus of Abdera and Manetho. Diodorus includes a long account of Sesoosis (another spelling of this pharaoh’s name), including the detail that he was sent into Arabia by his father with an army as his first military field experience. He conquered the entire nation of Arabs (1.53.5). Our fictional version does include a conversation between father and son, and what I just quoted clearly comes from a battle with the Arabs. However, in the third fragment—the encounter of Sesonchosis and the girl, Meameris—the non-Greek historical figure is transformed into an erotic hero. He is described as a good-looking youth (τὴν του̑ νεα[νί]σκ[ου] εὐείδειαν), and while he is in disguise in the court of the king of Arabia, the king’s daughter apparently falls in love with him. One fragment reads:
She reclined at the feast; she selected from what was placed before her with reluctance; she kept remembering the handsomeness of the young man (τὴν του̑ νεα[νί]σκ[ου] εὐείδειαν). And when she did not succeed in covering her feelings…. (3319 col. III 17–22)
It is the last section with its romantic overtones that places us in the realm of the Greek ideal novel. Whether our fragments belong to a full-fledged and early novel or a late local product, its interest is in the ways it blends the two cultures. Note the detail of the fleeing enemy—sumpateisthai—being trampled underfoot. This is surely a verbal statement of the traditional representations of the pharaoh routing the foe with enemy in topsy-turvy disarray crushed under the royal chariot or foot. Greeks were familiar with the trope: it occurs as early as the fifth century in vase representations of Busiris and Heracles (Boardman 1999), but here it appears not as parodic or ironic (as in the vases), but in its proper milieu. Sesonchosis, the pharaoh (or soon-to-be-pharaoh) is routing the foe of Egypt.
Iamblichus’ Babyloniaka has equally non-Greek affiliations, though if it had trans-cultural relatives they can no longer be identified. According to Photius (No. 94. 73b24–73b32), the Babyloniaka was less shameful in its events and language than that of Achilles Tatius, but more so than Heliodorus. The author (according to a scholium on Photius’ epitome) was Syrian by birth, not a Greek living in Syria, and he spoke Syrian. He learned Babylonian and about Babylonian culture from his tutor; he also claims to have written under Sohaimus, presumably the Syrian member of the Roman senate who was king of Armenia from 164 BCE (Stephens and Winkler 1995, 180–184). These explicit details of contemporary history, as opposed to the familiar retrojections into the Greek world of Herodotus and Xenophon, are unique and encourage a political reading of the text (as perhaps encoding resistance to Rome?). Certainly, the epitome indicates no Greek characters, but rather the hero and heroine are Sinonis and Rhodanes, a Babylonian couple either betrothed or actually married, who experience a series of adventures that makes the five canonical novels seem exercises in realism. Sinonis and Rhodanes are exact look-a-likes of Mesopotamia (Sinonis) and her twin brothers, Tigris and Euphrates (Rhodanes). There ensue mistaken identities, and curious marriages, for example, the king of Babylon wishes to marry Sinonis, hence tries to execute Rhodanes; the queen of Egypt, Berenice, “marries” Mesopotamia (see Morales 2006 for the political dynamics of this fragment); as the epitome closes, Sinonis agrees to marry the king of Syria. We also have poisonous bees, the ghost of a lustful billy goat, priests turned executioners, homicidal jealousies, battle scenes, and a finale in which Rhodanes defeats his rival the king of Syria, and apparently goes on to rule Babylon. Photius indicates that there was an excursus on magic practice, including discussion of various magi who worked respectively through locusts, lions, mice, serpents, and hail. He includes necromancy and ventriloquism among the magic arts. This kind of detail brings us closer to events in Apuleius and to the magician in Antonius Diogenes than to the five ideal novels. Our loss of this novel is substantial, because, like Antonius Diogenes, it seems to have been a sophisticated and complex narrative that differed from anything else that has survived, and thus would have provided further insight into novel-writing practices.
Iamblichus’ fictional characters operate in the same region as Ninus. According to pseudo-historical Greek accounts, Ninus was the eponymous founder of Ninevah, who was enamored of or married to Semiramis. Diodorus, Justin, and Strabo mention him, but our main Greek source seems to have been the fourth-century-BCE Ctesias, who is credited with a Persika, what was apparently a romanticized history of Assyria and Persia. Ctesias has not survived intact, but is epitomized by Photius (no. 72), immediately before Heliodorus. Our fragmentary Greek fiction alters this king, who was in earlier Greek texts a heroic warrior, and turns him into a lovesick boy, though a boy still capable of leading an army. The fragments that we have are similar to those of Sesonchosis: in fr. A, Ninus has returned victorious from his first military expedition and tries to persuade his cousin to move up the date for their marriage. In fr. B, Ninus is calming someone’s fears (probably his betrothed) and then goes off on an expedition against the Armenians; fr. C describes a shipwrecked Ninos on the Colchian shores. A late-second- or early-third-century-CE mosaic in Antioch depicts Ninus gazing at a picture of a girl, as a servant woman approaches with a cup. The event portraying both indicates the popularity of the fiction and lends credence to a reconstruction of the story in which Ninus is separated from his beloved (in good novelistic fashion). And it should be noted that no other novels besides Ninus and Metiochos and Parthenope so captured the imagination that they became subjects for visual depiction.
The Greek fragment presents us with a shy fiancée, who blushes to confess her affections for her betrothed even to her mother. The fact that the mother is named Derkeia makes it virtually certain that our fictional heroine would have been Semiramis, whose mother was the goddess Derketo in virtually all historical accounts. However, Semiramis was a warrior queen, not a retiring miss. The shift from a warrior queen to an untried girl, if it continued throughout the novel, runs counter to what we have come to expect from the canonical five, where the women are at least as enterprising as the men. Unlike Ninus, Semiramis seems to have had genuine Assyrian antecedents in Sammuramat, the wife of Shamshi-Adad V, who ruled Assyria from 823 to 810 BCE; at his death, she served as regent for her son. Sammuramat even managed to have her own deeds recorded in local inscriptions. Semiramis’ exploits received far more notice in Greek sources, in fact, than Ninus’, and, according to Stephanie Dalley,2 the erotic elements of the love of Ninus and Semiramis, which is found in Greek sources such as Diodorus, emerged rather late in the tradition and is to be seen in the figure of Stratonice in Lucian’s On the Syrian Goddess (Lightfoot 2004, 351–357).
If these four fragmentary novels have given us semi-historical and/or non-Greek protagonists, Antonius Diogenes gives us a narrative of interlocked tales that takes his characters out of the Mediterranean and known worlds entirely to Thule, a land that supposedly lay beyond the Outer Hebrides. Antonius was, according to Photius, “the father of similar fictions” (112a: ὁ τω̑ν τηλικούτων πλασμάτων πατὴρ), and Bowie (2002, 58–60) has given us cogent arguments to locate it in the mid-second century CE, if not earlier.3 Since it is one of the three novels Photius précised in some detail, it is likely to have been as rich and complex as Heliodorus—the only one of the three that has survived intact—and certainly the care Photius takes with explaining the series of interlocked narratives suggests a highly elaborate work.
The story as we have it falls into two discrete parts: the adventures of Deinias on the one hand and, on the other, of Derkyllis and her brother, whose name is probably Mantias, though Photius gives it as Mantinias, and P.Oxy. 70. 4760 as Mantias (see the discussion in that edition). Deinias and Derkyllis become lovers of sorts (what sort remains a question) after they meet on Thule, where she recounts her adventures to him. He then continues on beyond Thule, while she and her brother return to their home, Tyre, where Deinias later joins them. A central character in the Derkyllis portion of the story is an Egyptian priest, Paapis. Paapis has fled Egypt, when the country was in turmoil, perhaps around the time of the Persian occupation in the fourth century BCE, though Photius does not make this explicit. Paapis went to Tyre, where Derkyllis’ family befriended him. He apparently rewarded their philoxenia by persuading Derkyllis and Mant(in)ias to increase their parents’ longevity with a spell of his providing, but in the event it served only to cast them into a deathlike sleep. Our heroine and her brother were thus forced to flee Tyre, and as a result they had a series of increasingly improbable adventures. They again encountered Paapis, and this time they escaped with his satchel of magic spells and herbs (as it happens, one of our papyrus fragments, P.Oxy. 70. 4760, mentions Paapis and his satchel of books). Paapis catches up with them on Thule and turns them into zombies—dying each day and living by night—by spitting in their faces. Another character, Throuskanos, who has been in love with Derkyllis, in his outrage at this act, kills Paapis; then in his grief at Derkyllis’ seeming death, kills himself as well. The satchel of potions and spells are retrieved, and yet another character, Azoulis, uses them to discover how to reverse the spell first on Derkyllis and Mantinias, and later on their parents as well.
The name “Paapis” occurs in Manetho and Josephus, as that of the father of a famous sage, Amenophis, who was involved in the cleansing of the lepers from Egypt (fr. 54 Waddell). Our Paapis is surely not the same man, but the choice of a name with historical associations may have lent a semblance of plausibility to the fiction. It is difficult to gauge Paapis’ precise role in the novel. He obviously fascinated Photius, who reserves his final comments for him, noting that “even if he seemed to escape a thousand times he finally got his comeuppance” (112a 8–10), which suggests that he received even more narrative space than the current summary gives him. And given the tongue-in-cheek quality of Antonius Diogenes’ self presentation—he calls himself a poet of old comedy (111a 34–35)—it is quite possible that Paapis’ villainy (steeped as it was in exotic Egyptian sorcery and magic) was intended to provide yet another set of unbelievable things to entertain the reader.
One of the most striking features of this novel is the focus on its writing and preservation: when Deinias joins Derkyllis in Tyre, he has his adventures and all the many strange things he has learned recorded on cypress wood tablets. At his death, these tablets were buried near his tomb, later to be discovered, so we are told, by soldiers of Alexander at the time of the sack of Tyre. The rediscovery is framed by a letter from the author (Diogenes) to Faustinus, in which he apparently dedicates his work to Faustinus’ sister Isidora; he then quotes one Balagros’ letter to his wife Phila, which includes a transcript of the text that included the long history of events and copying of the narrative until it arrived in Deinias’ tomb. This is not a common feature in the surviving Greek novels, but we do find a similar fictional authentication in the tale of Dictys of Crete. In that text, Dictys asserts that he wrote his tale in Greek, but using Phoenician script. Writing on cypress wood bark, he too had it buried in his tomb. Many generations later, Cretan shepherds discovered the tomb and the enclosed tablets; they brought them to their master. The manuscript was then translated into Greek and sent to the emperor Nero in Rome. The prefatory letter in the Latin version is by a Lucius Septimus to Q. Aradius Rufinus. Septimius claims to have translated the Greek version into Latin (and in fact both Greek and Latin versions of this tale survive). The fact that these are unique frames suggests a connection, though whether one copied the other or both had a common model is moot. It is worth noting, though, that both are connected with Phoenicia—whether the city of Tyre as in Antonius Diogenes or in the script (as in Dictys of Crete)—and the closest parallel for copying texts, burying them in tombs, and rediscovering them, is to be found in Egyptian narrative, where it seems to be a staple; see, for example, the tale of Setne Khamwas and P.Oxy. 11. 1380, where an Egyptian priest discovers a sacred book long buried in a tomb. The proximity of Phoenicia to Egypt makes that country a potential source for the idea.
However, let us return to Antonius Diogenes. Another unusual aspect of this novel is the inclusion of Pythagorean lore, which seems to have been concentrated in Book 13 (there are 24 in total). Much of what has been preserved from this novel in Porphyry and John Lydus is owed to the fact that Antonius Diogenes apparently included a number of details about the life of Pythagoras and his teachings, packaging them in such a way that these subsequent (and serious) readers regarded the material as truthful. Heliodorus certainly has a neo-Platonic inclination, but Antonius Diogenes seems to have provided much more in the way of specific philosophical detail. Both John and Porphyry quote his discussion of Pythagoras’ dietary prohibition against the eating beans, and one character, Astraios, was said to have been a companion of the philosopher. Astraios’ eyes wax and wane in phase with the moon, a sign of his wondrous birth and nature (according to Porphyry, Life of Pythagoras § 10).
Antonius Diogenes’ novel differs from the canonical five in that the principal characters, Deinias and Derkyllis, seem not to be young lovers at all. They begin their adventures quite separately, and meet only on Thule, where they narrate their respective adventures to each other, separate, and finally reunite in Tyre. The years of age inscribed on their respective tombs suggest at best that Deinias was a man in mid-life, and nothing in what remains suggests teen lovers. Since all of the other fragments we have discussed seem to associate eros and youth, The Unbelievable Things beyond Thule looks like a radical departure (or, if it was early enough, a novel written before the pattern took hold). Other elements of the narrative—Egyptian magic in the person of Paapis and Pythagoreanism—align the novel in some respects with Apuleius’ Metamorphoses, where an Isiac redemption is tacked on to a tale of magic misadventure, but not enough of the text survives to press the parallels much further.
The final text of any length is Lollianus’ Phoinikika. Three fragments of this novel, all from the late second and third centuries CE, have been published to date: the longest is from a second-century codex that contains two tantalizing incidents: (1) a man experiences his first sexual intercourse, and (2) what seems to be the ritual murder of a pais (either a servant or child) by outlaws who proceed to cook and eat their victim. When the fragment was first published in 1969, the sensational nature of these events quickly caught scholarly imagination, and discussions of their meaning ranged from confirmation of the view that the novels were coded descriptions of ritual events in mystery religions (e.g. Merkelbach 1962) to interpreting it to be a narrative deception parallel to the Scheintod found in Achilles Tatius. There are numerous characters, but only one, Glaucetes, links the three fragments. He seems to be a servant. In the codex fragment, he has a mistress who instructs him to bring her 2,000 drachmas from her steward. In the second fragment (P.Oxy. 11. 1368), Glaucetes, en route to an unknown destination, encounters a ghost who tells him: “I am lying there beneath that plane tree, and with me a fair maiden, both of us slain” (3–5). In the third fragment (P.Oxy. 73. 4945), we find characters burning with love and Glaucetes instructed to arrange an assignation. Unfortunately, it is not possible to reconstruct an overarching narrative from what survives, nor to discern whether the novel is like the canonical five with a Liebespaar who experience grave dangers, one of which is the ritual murder (or Scheintod), or if we have a novel closer in style to Apuleius’ Metamorphoses. It is worth noting that, if we only had the two fragments from papyrus rolls, we would evaluate the remains quite differently than we must now do in light of the long codex piece.
We know the name of the author and the work from a subscription on one of the fragments, and although it is tempting to identify him with a sophist (Lollianus the Ephesian) who wrote under the name “Hadrian” according to the Suda, there is no compelling reason to believe they are the same, since Lollianus was a relatively common name. The novel’s title, Phoinikika, links it with works such as Iamblichus’ Babyloniaka or Heliodorus’ Ephesiaka, but also with pseudo-historical works such as Ctesias’ Persika. However, as Tim Whitmarsh (2005, 601–603) has demonstrated, these names, which are “content descriptors,” are sufficiently generalized that they do not guarantee one or another type of plot.
The narrative trajectory of events in the codex fragment and that of Lucius’ captivity in the robbers’ cave (Apuleius 4.8 and 22) are sufficiently similar in detail that they suggest some textual relationship (see, e.g. Stephens and Winkler 1995, 322–325). Did one borrow from the other? The dating of the papyrus fragments does not aid in establishing which direction the influences may have occurred, since the Phoinikika could have been written before Apuleius or after. A more likely scenario is that both were dependent on the original Greek ass-tale that Photius epitomized but which has not survived (Photius no. 129 attributes it to “Loukios of Patrai” who sounds like the narrator of the story, though he may have been the author of an ego-narrative). This is not a particularly bold claim; just as the canonical five seem to have intertextual connections, we should expect to see, as these fictions evolved and proliferated, that they would build on earlier narrative patterns. In addition to the lost ass-tale, an epitome of that story, falsely attributed to Lucian, now exists, and a newly published papyrus scrap (P.Oxy. 70. 4762) relates an encounter of a woman and a donkey which suggests that it too is a version of the ass-tale. Thus, we have a circulation wide enough to accommodate several different levels of narrative sophistication.
If Lollianus has affinities with Apuleius, another very small scrap, the Iolaos fragment, has affinities with Petronius, and was originally published under the title “A Greek Satyricon?” (Parsons 1971). In this fragment, a male character seems to try to gain access to a girl by pretending to be one of the galli, the castrated priests of Cybele. It combines prose and verse with a central run of Sotadeans set between two sections written in prose. The final lines of the column are an adaptation of Euripides’ Orestes 1155–1157. The affinity with Petronius depends on the salacious aspect of Iolaos’ plot in combination with prosimetrum. Greek Menippean satire was known to have been prosimetric and salacious, and to have been influential on Varro, who imitates it, and especially on Petronius (see Baldwin’s essay in this collection). The relative size of the fragment (remnants of only 43 lines) makes generalizing dangerous. Still, the existence of this kind of narrative in Greek in Egypt in the second century CE suggests (as with Lollianos) a rather wide dissemination of this kind of fiction (whatever its length) within Greek-speaking communities of the early empire.
In addition to these fragments, at least four others exist that have some claim to novel status: Herpyllis, Kalligone, Antheia, and Chione. In contrast to the two previous fragments, in each of these the central figure is easily recognizable, and a woman. All of them detail what appear to be serious incidents coincident with the ideal novels, either of adventure or love and marriage. The last three, to the extent we can judge, resemble Chariton in their straightforward narrative style. Herpyllis consists of one broken and one nearly intact column, the latter of 60 lines. It recounts the departure from land into the Cretan sea of a couple (whether brother–sister or Liebespaar is not known) on two separate ships. The male narrator finds himself separated from Herpyllis as his ship is engulfed in a fearsome storm when the fragment breaks off. Affinities with the canonical five are obvious, though the couple on separate ships is an uncharacteristic detail. The description of the storm is a rhetorically elaborated set piece consonant with the best surviving fictions. The woman’s name is telling: it is rare and closely resembles the name of Antonius Diogenes’ heroine, Herpyllis, rather as Antheia’s name in Xenophon of Ephesus resembles Pantheia’s in the Cyropaideia. However, the nature of the intertextual relationship (if any) remains a matter of speculation.
Kalligone gives us a glimpse of a passionate heroine: she stalks into a tent in fury, throws herself on the bed, and rips her clothing in mourning. The location is the Bosporus, in the territory of the Sauromatians and Amazons. Why she is in such dire distress is not clear, though one may suspect that she has heard that her beloved has been killed. One of the characters named Eubiotos was, according to Lucian’s Toxaris, an illegitimate brother of the king of the Bosporans. He lived among the Sauromatians and at his brother’s death became king (§ 54). There are narrative overlaps with Lucian as well, which led M. Rostovtzeff (1931, 98–99) to argue for the existence of “Scytho-Bosporan” romances. Certainly, this fiction has a historical character, like Chariton or Metiochus and Parthenope, but its relationship to Lucian’s text should make us aware that such writing need not be a novel but a lively incident in a work that imagines itself as history. However, if we consider Bosporan Kalligone, Egyptian Sesonchosis, Babylonian Ninus, and the Greek Metiochus and Parthenope, it appears as if there was a strand of Greek novel or novel-like writing that depended on, or was developed out of, historical characters and circumstances. Chariton fell into this category, but the other four Greek novels have very attenuated connections to real events or earlier historical figures.
Chione is a palimpsest fragment that comes from a codex that contained the ends of Chariton’s Callirhoe, so prima facie Chione is also a novel (and may even be by the same author, though there is no way to test this hypothesis). Unfortunately, the codex itself was lost after it was partially transcribed. The surviving plot revolves around Chione, the daughter of a king, who has promised her to a powerful warlord named Megamedes. For some reason, the king no longer wishes to honor his commitment, possibly because Chione has fallen in love with someone else, and Megamedes with his army has issued an ultimatum: honor the commitment or he will invade the city in 30 days.
Antheia is an even smaller scrap but with intriguing parallels to Xenophon’s Ephesiaka. We have: a young woman (Antheia?) concealing poison about her person, apparently prepared to commit suicide “lest she be captured (?) again”; mention of an abduction; characters with names that also occur in Xenophon (Antheia, Euxeinos); and apparently a temple of Artemis (and it was in a procession of Artemis of Ephesos that Xenophon’s lovers first met). Since many have suspected that the current version of the Ephesiaka was an epitome of a longer narrative, the discovery of this scrap raised the question of whether it might belong to an earlier or fuller version of Xenophon’s novel. Apart from wishful thinking, there is no reason to believe that the fragment and Xenophon’s novel are so closely related.
What then can we say about the ancient novel on the basis of these fragments? There are several conclusions to be drawn: (1) the five extant Greek novels are very similar in their plot lines and may well represent (as Bowie and Whitmarsh would argue) a subset of interrelated texts that came to dominate (and subsequently displace) other types of narrative. Young women as protagonists are found in all but Lollianus. (2) Novels with an erotic focus that nevertheless had strong connections to historical figures, such as Ninus, Sesonchosis, or Eubiotos, were also popular, though they seem to have mainly disappeared. The reason for this may have been that they were too closely connected to one or another region to have survived long past the time of their composition. (3) As we have seen from this all-too-rapid survey, all of the fragments that survive seem to display some connection to other extant novels (or to each other). This suggests that novel writing fairly quickly developed a set of recognizable characteristics, and this perhaps (like the genre novels familiar to us today) served to guarantee a certain level of audience interest. Of the four novels that Photius epitomizes, two of the three to which he devotes the greatest space have disappeared. Since they seem not to have conformed to the pattern of the five surviving novels, although from their epitomes and fragments they appear to have been of equal quality, it would appear that selection rather than happenstance eliminated them from the field. Their loss, more than the others, limits our understanding of fiction writing, since it eliminates texts that appear to have been both ambitious and atypical. (4) The astonishing afterlife of Parthenope (with or without Metiochus) is a sobering reminder that the Greek and Roman novels should not be treated as a closed set—Christian martyr acts have strong antecedents in the novels—and relationships with the narratives of other cultures (Egypt, Persia, Syria, Judea) are equally important in grappling with questions of growth, development, and influence.4
Finally, to return to my opening remarks, identifying and discussing fragments, like the bounding of the field of the novel, is inherently circular in its processes. What necessarily falls out of my assessment is any fragment that, for reasons of length, does not display certain characteristics. Hence, fragments now believed to be history, magic, religion, rhetorical exercise, or even mime might, with the addition of two columns, look entirely different, as what currently passes for salient generic markers are subsumed into new and considerably more elaborate narrative strategies.
1 For up-to-date lists of literary papyri, see the Leuven Database of Ancient Books (http://www.trismegistos.org/ldab/) and CEDOPAL (http://www2.ulg.ac.be/facphl/services/cedopal/pages/mertensanglais.htm).
2 http://www.classics.ox.ac.uk/romance/workshop2.htm. Accessed 30 June 2010.
3 A late first or early second century date means either Photius was mistaken or (more likely) that he considered Antonius Diogenes as the earliest of the four novels that he epitomized.
4 A series of workshops on the “Romance between Greece with the East” held at Oxford in 2009 make this abundantly clear.
Boardman, J. 1999. The Greeks Overseas: Their Early Colonies and Trade. London and New York: Thames and Hudson.
Bowie, E.L. 2002. “The chronology of the earlier Greek novels since B.E. Perry: Revisions and precisions.” Ancient Narrative, 2: 47–63.
Bowie, E.L. 2003. “The ancient readers of the Greek novels.” In The Novel in the Ancient World (rev. ed.), edited by G.L. Schmeling. Leiden: Brill, pp. 83–113.
Hägg, T. and B. Utas. 2003. The Virgin and her Lover: Fragments of an Ancient Greek Novel and a Persian Epic Poem. Leiden: Brill.
Kussl, R. 1991. Papyrusfragmente griechischer Romane. Ausgewählte Untersuchungen. Tübingen: G. Narr.
Lightfoot, J.L. 2004. Lucian: On the Syrian Goddess. Oxford: Oxford University Press.
López Martínez, M.P. 1994. Fragmentos papiráceos de novela griega. Alicante: Universidad de Alicante.
Merkelbach, R. 1962. Roman und Mysterium in der Antike. Munich and Berlin: Beck.
Morales, H. 2006. “Marrying Mesopotamia: Female sexuality and cultural resistance in Iamblichus’ Babylonian Tales.” Ramus, 35.1: 78–101.
Morgan, J.R. 1998. “On the fringes of the canon: Work on the fragments of ancient Greek fiction 1936–1994.” Aufstieg und Niedergang der römischen Welt, II:34.4: 3293–3390.
Parsons, P.J. 1971. “A Greek Satyricon?” Bulletin of the Institute of Classical Studies, 18: 53–68.
Rostovtzeff, M.I. 1931. Skythien und der Bosporus I. Berlin: H. Schoetz.
Stephens, S.A. 1994. “Who read ancient novels?” In The Search for the Ancient Novel, edited by J. Tatum. Baltimore and London: Johns Hopkins University Press, pp. 405–418.
Stephens, S.A. and J.J. Winkler 1995. Ancient Greek Novels: The Fragments. Princeton: Princeton University Press.
Stramaglia, A. 2000. Erôs. Antiche trame greche d’amore. Bari: Levante.
Whitmarsh, T. 2005 “The Greek novel: Title and genre.” American Journal of Philology, 126: 587–601.
Widmer, G. 2002. “Pharaoh Maâ-Rê, Pharaoh Amenemhat and Sesostris: Three figures from Egypt’s past as seen in sources of the Graeco-Roman period.” In Acts of the Seventh International Conference of Demotic Studies 1999: 387–393.
S.A. Stephens and J.J. Winkler, Ancient Novels: the Fragments (Princeton 1995), provides the most accessible introduction to fragmentary novels. Since its publication, new fragments of Lollianus, Antonius Diogenes, and Dictys of Crete have appeared in the Oxyrhynchus Papyri, vols. 70 (2006), 71 (2007), and 73 (2009). On Lollianus, J.J. Winkler’s “Lollianus and the Desperadoes,” JHS 100 (1980) 155–181 provides an elegant model for interpreting fragments. T. Hägg and B. Utas’ The Virgin and the Lover: Fragments of an Ancient Greek Novel and a Persian Epic Poem (Leiden, 2003) is an excellent study of the transmission of Metiochus and Parthenope from one culture to another.