20. Ptolemaic Homers

We may correlate this period of innovation in rhapsodic performances, roughly from the fourth to the first centuries BCE, with a small corpus of Homeric papyri from the Ptolemaic period (305–145 BCE) that stand out for the peculiarity of their divergences from the vulgate of Homer. The Ptolemaic papyri of Homer, [1] collected and edited by Stephanie West, give us many examples of so-called “plus-verses,” which are additional verses that survive but do not appear in the vulgate Homer as it becomes standardized after the editorial activity of Aristarchus, perhaps in 150 BCE or so. These papyri, dating from about 300 to 150 BCE, are considered “eccentric” or “wild” (terms that are not really appropriate [2] ) because they diverge so much from the Roman papyri of Homer, which deviate much less from the medieval manuscripts. As Stephanie West points out, these papyri “cannot be explained by the processes of mere mechanical (that is to say, scribal or copyist) corruption.” [3] The divergences simply show too intimate a knowledge of the Homeric texts to be “errors” in the usual sense, and are more readily understandable as the product of a still lively poetic tradition. [4]

Many experts including Thomas Allen have argued that these variations are specifically due to the performance of rhapsodes. [5] This same conjecture was made in the nineteenth century (in the wake of Friedrich August Wolf’s rhapsodic Liedertheorie of the composition of the Iliad and Odyssey), but at that time scholars like Arthur Ludwich regarded rhapsodes such as Cynaethus as inferior forgers and falsifiers of the Homeric text. [6] Allen adopted this same prejudice when, following Ludwich, he argued that rhapsodes were attempting to “increase and improve” the Master, [7] that is Homer, whence he proceeded to give an allusion to Mozart’s supplements to Handel. More recently, Michael Apthorp has argued along similar lines that the Ptolemaic papyri should be understood as “lapses of memory” or the result of inevitable “alterations and additions to the poems in the process of recitation” by rhapsodes that arise during an oral performance. [8] But such a view flies in the face of the clearly detailed knowledge of Homeric diction and formulae in the variations. To the contrary, it is more likely that these papyri reflect different ground rules for competitive performance. The variations often reflect deliberate rearrangement of formulae and verses to achieve unexpected effects, which would have appeared to an audience unprepared for such changes as refreshing novelties. A similar position is maintained by Aldo di Luzio, whose study of the Ptolemaic papyri shows that the variations are not random but fall into recognizable patterns with poetic and rhetorical functions. [9] Some of the Ptolemaic variations (e.g. Iliad 5.808) actually reflect a pre-Panathenaic repertoire that Aristarchus refrained from including based on the manuscript evidence at his disposal, but which then reappear in post-Aristarchean manuscripts of the “vulgate.” [10] Thus earlier and later phases of Homeric transmission were fair game. At the same time, the Ptolemaic papyri appear during the very period in which other types of innovation in rhapsodic performances in the Greek-speaking world emerge, including the creation of new epic material. So it is more pertinent to ask why some Greek-speakers in Egypt preferred, at least as indicated by the eccentric papyri, to accept a reorganized text of Homer in part reflecting earlier traditions—some predating the Panathenaia—rather than to create entirely new material. Their actions reflect very specific performance demands in addition to, as others have argued, [11] a generalized reintroduction of fluidity into the textual tradition.

In this connection it is worth noting two related details about rhapsodes that involve the manipulation of Homeric material within individual verses. These examples give added dimension to the potential subtlety of their performances. The first involves an anecdote in Plutarch about Ptolemy II Philadelphus on his wedding day. [12] Ptolemy II married his sister Arsinoe, who would become one of the most important women rulers in Egypt, yet at the time the marriage was considered scandalous by Greeks. Plutarch relates the story of the rhapsode whom Ptolemy II had hired to perform at his wedding—a not uncommon practice [13] —and this rhapsode became famous for beginning his performance with line 18.356 of the Iliad:

Whoever this rhapsode was, he was clever enough to begin his performance by adducing an apt line from Homer, but there is greater subtlety to his recitation than has been noticed. In the vulgate of Homer, this line does not say that Zeus summoned (καλέω) Hera, with its more stately implication, but rather the following (Iliad 18.356):

In other words, according to the vulgate Zeus merely spoke to or addressed (προσεείπω) Hera at this point, and what follows this line is actually a speech by Zeus. Although we do not know the source of Plutarch’s quotation, it is possible that our rhapsode not only aptly quoted this line of Homer, but also that he improvised the verb—whether it was drawn from scratch or an earlier tradition is not clear—to make the whole line more consonant with the circumstances of Ptolemy’s wedding.

The second example comes from the T scholia to Iliad 21.26. After a description of Achilles’ slaughter of Trojans in the Xanthus river, he wearies of killing and then captures twelve Trojan youths as a recompense for the dead Patroclus. Of Achilles’ fatigue specifically, we read (Iliad 21.26): The idiom in Greek requires that the noun χεῖρας, in the accusative, represent the body part that is fatigued in connection with the verb κάμνω ‘to weary’, while the participle ἐναίρων (from ἐναίρω ‘to slay, kill’) describes the action from which one is fatigued. However, the T scholia report that a rhapsode named Hermodorus (otherwise unknown) placed a different construction on this line. The scholion reads: Ἑρμόδωρος ὁ ῥαψῳδὸς χεῖρας ἐναίρων ἤκουε χειροκοπῶν, κατεχρήσατο δέ. “The rhapsode Hermodorus for χεῖρας ἐναίρων heard ‘hand-cutting’, and used it wrongly.”

This is a case where it is important to distinguish between Hermodorus’ interpretation of the line and the scholiast’s condemnation of his syntactic knowledge. If our bias is against the ingenuity of rhapsodes, the scholiast’s judgment will seem correct. But if we allow that rhapsodes were capable of clever improvisation, then we may detect a hint of deliberate playfulness in Hermodorus’ interpretation of Achilles’ actions at this point in the narrative. By taking the noun χεῖρας as the direct object of the participle ἐναίρων, rather than with κάμνω, Hermodorus represents Achilles as actually cutting off the hands of the twelve youths whom he will capture in the following lines as recompense for Patroclus (21.27–8). Rather than misunderstanding or misapplying Greek syntax, I interpret Hermodorus’ play as a purposive improvisation meant to depict Achilles in a more gruesome fashion. Not all such examples, of course, need to be interpreted as deliberate. There were famous examples of performance gaffes in antiquity, remembered for their malapropism, such as the actor Hegelochus’ mispronunciation of γαλήν᾽ ὁρῶ ‘I see calm [weather]’ as γαλῆν ὁρῶ ‘Ι see a cat’ in a performance of Euripides’ Orestes (line 279). This gaffe was frequently cited especially by comic poets. [14]

But as in the previous example of Hermodorus, we must be cautious not to assume that Alexandrian scholiasts had a model of performance of Homer before them as we do. Whatever they witnessed in performance is lost to us. Yet when it came time to edit Homer, they were concerned with noting what they perceived to be contamination in the manuscripts at their disposal, though it is not always clear what their manuscript base was. [15] What might have seemed to them irrelevant or wrong from the standpoint of textual criticism, might be the very types of changes that offer direct evidence of performance variations characteristic of rhapsodes. The fact that Hermodorus’ interpretation is reported at all suggests that his violation of Greek syntax nevertheless resulted in a striking and memorable image.

Rhapsodes alone may not be the only performers responsible for textual changes or improvised interpretations of Homeric poetry. We must also briefly consider the figure of the ὁμηριστής (Latin homerista), about whom much less is known but who is closely related to the rhapsode and mime. [16] Let us take the connection with rhapsodes first. In at least one account, the homêristês is actually said to be identical with the rhapsode (Athenaeus 620b). There has been some dispute over the exact historical relationship between rhapsodes and homêristai, because the name homêristês, derived from the verb homêrizein ‘to act Homer’, [17] in other contexts suggests that they both recited and mimed Homeric poetry. [18] Nevertheless, the connection between them seems to relate to the degree of acting involved in the performance of Homer, with the homêristai representing a more dramatic phase in the tradition.

In the third quarter of the fourth century, when Demetrius of Phalerum (ruled 317–307 BCE) was at the height of his political and cultural influence in Athens, we are told that he was the first to introduce those who are called homêristai into the theaters: I cite this passage at length because it provides significant background to the wide variety of poetry that was performed in theatres, such as the hexameters of Hesiod and Empedocles, [19] and also the iambic poems of Archilochus and Simonides. Most importantly for the present, however, is that the great theatre of Alexandria is singled out as the locale for the acting (ὑποκρίνομαι) of Hesiod and Homer. To follow Athenaeus’ logic of presentation, even the fact that the poetry of Homer and Hesiod was acted by a comedian (κωμῳδός) in Alexandria can be seen as a development of the greater theatricalization of Homeric performance begun by Demetrius.

Athenaeus says explicitly that the term homêristai was another name given to rhapsodes, hence our need to confront the homêristai more directly. In general our evidence for the nature of their performances is very scant, but other literary evidence in conjunction with several papyri suggest that both in large-scale public and smaller-scale private venues homêristai performed well into the third century CE. So for example, one incidental reference to what the homêristai did comes to us from the Interpretation of Dreams by Artemidorus, dated to the second century CE, in which there is an anecdote about a surgeon who once dreamed that he was acting Homer. The surgeon draws an analogy between the motions made by homêristai as they gesture in performance and those made by a surgeon as he operates (Artemidorus 4.2, ed. Pack): [20]

Another passage from Petronius’ Satyricon is more descriptive, but also gives some indication of the changing venues for homêristai performances. In this passage, Trimalchio, a poorly educated but wealthy freedman who is in the midst of feasting his friends at his home, asks that everyone be festive and watch the homêristai as they make their entrance:

There is much humor in this scene—of course not only are the homêristai lavishly decked out in military armor but their dialogue is loud and affected. Moreover, Trimalchio obviously knows no Greek and therefore must read along in his Latin translation of Homer to follow the performance. [21] Trimalchio becomes more of a fool in what follows, when he asks the homêristai to stop while he explains the plot to them. He completely confuses the characters by saying that the brothers concerned were Diomedes and Ganymede (instead of Agamemnon and Menelaos); that their sister was Helen, whom Agamemnon rescued and then substituted a deer in her place, as an offering for Diana. He goes on to say that Agamemnon gave his own daughter Iphigeneia as a wife to Achilles, and that on account of this (instead of Achilles’ armor) Ajax went insane (59.4–6). This is all quite absurd, but finally, at the mention of Ajax, Trimalchio’s servants begin to scurry about making preparations for the entry of a boiled calf, which is brought in on a heavy tray with a helmet on its head. Then a man dressed as Ajax, possibly a homêristês, comes in with a sword and begins to mime as if he were the insane Ajax madly cleaving at herds of cattle, all the while collecting bits of meat on the end of his sword and passing it to the guests who look on in amazement (59.6–7). For our purposes, this parodic display does at least support the idea that the homêristai, who not only performed in theatres but as we have just seen could also be hired out for élite dinner parties, both recited Homeric verses and mimed the dramatic action. [22]

Other evidence for homêristai performances comes from papyri dating from the second to the third centuries BCE. These papyri are contemporary with the eccentric papyri of Homer and show connections with mimes. As the papyrologist Genevieve Husson has demonstrated, there are at least five papyri from Oxyrhynchus, some of which are contracts for actual performances (with fees indicated), in which homêristai are sometimes paired with mimes. This may suggest that the homêristai recited Homer while the mimes did the acting. [23] More recently Michael Hillgruber has argued that homêristai are better considered a type of mime that specializes in performing Homeric material rather than an altogether different kind of performer. [24] In any case, the setting would once again have been of large-scale public performances like that of rhapsodes—we know for example that the theatre at Oxyrhynchus could hold upwards of 11,000 people [25] —and the context for these performances would have been competitive. Indeed one papyrus, P. Oslo 3.189.19 studied by Husson (her text 2), mentions a contest of poets (ἀγὼν ποιητῶν), somewhat along the lines of the Hellenistic victory inscriptions discussed earlier.

Taken together, then, this evidence for homêristai suggests that, by virtue of their need to recite Homer in performance, they too could be responsible for the variations that we find in the eccentric papyri. Rhapsodes are rarely mentioned in the Homeric scholia, [26] but this may have nothing to do with whether they created or annotated copies of Homer for themselves, as is occasionally attested for schoolteachers. [27] From the standpoint of trying to explain the Ptolemaic textual variation, we cannot exclude other performers of Homer like the homêristai, the content of whose performances largely elude us but which could have demanded the special effects achieved in the eccentric papyri. [28]

To restate the argument briefly: the evidence we have for rhapsodic performance suggests that rhapsodes competitively recited memorized verses, improvised verses or portions of verses on the spot for elaboration or embellishment, and took up and left off Homeric (or other) narrative wherever they chose. Further evidence suggests that rhapsodes modified words within a verse, or modified Greek syntax where plausible to create new meaning from a known verse. To the extent that homêristai performed in a manner comparable to rhapsodes, we may attribute similar skills to them, although it does seem that homêristai were more theatrical. Viewed in this light, the Ptolemaic eccentric papyri may show direct evidence of this kind of manipulation. What we now need to explain are the effects achieved by some of the “plus-verses,” which are the distinguishing feature of these papyri.

The creation of a vivid and memorable image is a case in point. A typical example comes from Iliad 22, line 316, in the scene where Achilles lunges at Hector. This is one of several performable scenes or episodes mentioned by Socrates, in Plato’s Ion (535b). In the Iliad scene, Hector and Achilles have exchanged some boasts and abuse, and then Hector calls upon his brother Deïphobos to give him a spear. Realizing that Deïphobos is not close enough to do this, Hector senses that his fate is near, and so gathers himself together and makes a run for Achilles. At this moment, Achilles charges in return, and we read about his helmet glittering in the sun, with its golden plumes (Iliad 22.316):

In P12, a papyrus datable to the early to mid third century, we find three plus verses to accompany line 316, [29] which I give below (P12 Iliad 22.316a, b, c = 22.133–35): Note especially that the enjambing word (σείων) in 316α is a participle, a frequent and flexible type of enjambement in Homer and the Certamen, and that the participle is consistent with the uses of enjambement by rhapsodes for which I argued earlier. In any case, these three verses are identical to verses 133–35 from the same book, 22, of the Iliad, as transmitted through the vulgate. Now the question is, why do these plus verses appear at line 316 in this eccentric papyrus?

I think we can provide an answer, but in order to do so we have also to supply a little imagination. We have to suppose that our audience knows book 22 well enough to know the narrative context of lines 133–35, and that the performance venue for these verses is a rhapsodic or homeristic contest. Before the lines occur, King Priam and Queen Hecabe have unsuccessfully attempted to keep Hector from battling Achilles. Hector then reflects on the tight position that he is in: if he retreats he will be ridiculed, but since he has by his own recklessness endangered the Trojans, he feels compelled to continue fighting. He then debates with himself about refusing to fight, giving up Helen, and even laying down his armor and propitiating Achilles. This does not seem satisfactory either, and so he resolves to let Zeus decide the victor. It is at this moment that we see Achilles closing in on Hector, shaking his dangerous Pelian ash spear by his right shoulder with his helmet blazing in the sun. So run verses 22.133–35. Now when Hector sees this, he can no longer stand his ground and so flees, frightened, toward the base of the Trojan wall. Clearly the appearance and description of Achilles is decisive for Hector at this moment, yet it is not until Achilles’ next lunge for Hector, at lines 312 and following, with our plus verses in the papyrus, that he will make the fatal spearthrust through Hector’s throat.

Therefore I am suggesting that, given our hypothetical performance context, lines 316a–c could be an “improvisation” on the part of one rhapsode or homêristês who is simply embellishing and intensifying the description of Achilles at the fatal moment for Hector. I use “improvisation” in quotes because these verses might have been composed beforehand and then presented in performance, only to seem freshly improvised from the audience’s point of view. They are not necessarily an actual improvisation that was later recorded. In any event, for an audience who knows their Homer, the verses add more vividness to the description of Achilles’ final lunge at verse 312 and following. Of course we cannot say whether a rhapsode embellished line 316 as a virtuoso flourish, or whether a homêristês employed the lines parodically to accentuate the presentation of a costumed mime, impersonating Achilles, as he stood there brandishing a spear in defiance. Both scenarios are equally plausible.

One consideration of great interest is whether verses from the Iliad could be transposed to the Odyssey, or vice versa, or whether the variations had to be manipulated within one epic alone. Because we have certain evidence that verses were transposed between epics, this tells us that there was a large degree of freedom available to performers and it also indicates that a wide range of effects could be achieved through such manipulation. For example, a formulaic phrase brought from the Iliad into the Odyssey has the potential of evoking contextual differences that create “new” meaning. A well-known example comes from P30, a papyrus datable to the second century BCE, which has remains of lines from Odyssey books 4 and 5. [30] On the morning of Odysseus’ departure from Calypso’s island, both the hero and goddess dress themselves. For Calypso, the vulgate has (Odyssey 5.230–32):

After these lines in P30 occur two more lines, which Ι reproduce without brackets and sublinear dots because the attestation is certain (5.232α–β): These additional lines have caused much consternation, beginning with whether the καλύπτρη ‘veil’ was to be identified with the κρήδεμνον ‘headdress’ or whether these terms ought to remain distinct. [32] Because the lines describe a second veiling of Calypso, they have been viewed more recently by some commentators as “incoherent.” [33] But to take such a position is to assume a narrative standard that does not reflect the performance demands of Alexandrian rhapsodes, homêristai, or whoever performed these texts. As di Luzio has cautioned, we simply cannot apply modern standards to the Ptolemaic papyri, especially those derived from the study of written literature; [34] the variations reflect different performance objectives. Line 232α and the first hemistich of 232β are found in Iliad 14.184–5, in the context of Hera’s toilette in anticipation of her deceptive tryst with Zeus. The only difference is that the second hemistich of Iliad 14.185 reads λευκὸν δ᾽ ἦν ἡέλιος ὥς “it was as bright as the sun,” while the phrase in the papyrus, τό ῥά οἱ τεθυωμένον ἦεν “which was fragrant for her,” occurs several lines earlier at 14.172 and refers to the sweet smell of the olive oil with which Hera anoints herself. We have, then, formulaic lines transposed from a similar situation and added on through enjambement, which is a typical mark of rhapsodic manipulation. [35]

But similarity of narrative context cannot be the only reason for the transposition. A closer consideration of the additional lines suggests that their author intended to evoke the sexual atmosphere of Hera’s toilette in the aftermath of Calypso’s encounter with Odysseus. In a few lines preceding the description of Calypso’s dressing, we are told at Odyssey 5.227 that on the previous evening Calypso and Odysseus had slept together. It is reasonable to suppose that after the lines describing Calypso’s veiling, the author of P30 chose subtly to transform the description of the veil by adding details drawn from an explicitly sexually charged situation in the Iliad. A further nuance is achieved by P30’s transposition of the hemistich, τό ῥά οἱ τεθυωμένον ἦεν, which adds fragrance to the κρήδεμνον. If this is the intended effect of 232α–β, it is then beside the point to speculate about whether the καλύπτρη is identical with the κρήδεμνον, or in which order the two have to be put on to retain “coherence.” [36] The more important point is that the author of P30 intends for Calypso’s veil to be sexually attractive to Odysseus, and possibly further to suggest that Calypso intends to deceive him along the lines of Hera’s deception of Zeus in Iliad 14. What this example shows is that an intimate knowledge of epic formulae as well as of contextual details figured in the choice of variations. Lines were manipulated in performance but not incoherently. We cannot say for certain how attentive to the details of these effects a Ptolemaic audience might have been. But if we only apply modern standards of coherence, we run the risk of obscuring the sophisticated ways in which the variations create narrative effects.

In all this I am not insisting that we always assume a one-to-one correspondence between a given papyrus and a given performance, or that these texts are necessarily scripts or memory-aides for performance. This hypothesis goes back to Kirchoff in the nineteenth century and, while it remains attractive, we still do not know the true origin of these papyri. [37] However, the advantage of the approach outlined here is that it offers an alternative to attributing such plus-verses and variations to pedantic scribes or misinformed copyists, or to dismissing them as uncreative interpolations of inferior performers. The question of the authenticity of these variations is also misleading to the extent that it presupposes an “original” text against which the variations ought to be judged. Instead, it makes more sense to view the variations as stemming from or produced for live performances during which variations were expected—which also means that to this extent they must all be regarded as “authentic.” [38] The variations suggest that knowledge of Homeric texts and an ability to manipulate passages were of primary importance to the authors and performers of these papyri, because the innovation here involved the novel deployment of traditional material within and between the Iliad and Odyssey. But the motivation for the variations is best explained by the competitive context of rhapsodic performances or, possibly, the parodic context of homeristic performances, although the “stitching” nature of the variations in the papyri inclines me toward a rhapsodic performance scenario. Competition produced the demand for “new” narrative effects, striking images, and unusual juxtapositions, by performers intimately familiar with the Homeric text and its history. What we can probably exclude is the possibility that the variations were due to poets, [39] because as we saw earlier in the discussion of Hellenistic performances from the fourth to the first centuries BCE, the so-called poets of epic (ποιητὴς ἐπῶν, ἐποποιός) typically were rewarded for the creation of new epic material largely treating historical and mythological subjects. [40] We may conclude that these papyri reflect the interests of a delimited group of performers/authors who specialized in Homer, because we do not find the same extent of verse manipulation in Homeric papyri after 150 BCE, even though rhapsodic (and homeristic) performances continued until the third century CE. For this reason, I regard it as more than probable that these papyri issued from the Ptolemaic equivalent of the Homeridae of Chios or the Creophylei of Samos.

A comparison with the performance of tragic poetry may also be instructive here. We know that by 330 the Athenian statesman Lycurgus, as part of his major theatrical reforms, [41] apparently sought to curtail the improvisations of actors with a decree limiting their lines to fixed texts of Aeschylus, Sophocles, and Euripides ([Plutarch], Lives of the Ten Orators 841f). [42] But the details of this decree are worth considering more closely because this view assumes that the texts of the tragedians were in a state of deterioration and needed to be rectified. Lycurgus first ordered that bronze statues of the three tragedians were to be erected, and then enjoined:

The problems with the last sentence, as with the passage as a whole, have a long scholarly history. [44] Arguing from classical period usage, Bollack suggests that παραναγινώσκειν means that the secretary of the polis ‘collates and checks’ available manuscripts for the actors, rather than ‘reads alongside’ them so as to regulate their performance. [45] Further, he argues that the decree itself offers no hint that the texts of the tragedians had undergone any deterioration, as if through unauthorized actors’ interpolations or scribal errors. [46] Rather, it is more likely that actors’ scripts with theatrical annotations as well as textual editions for private reading provided the disparate material from which the polis secretary was to produce a more “original” version. [47] The advantage of this perspective is that it allows for textual variations due to performance to have become by Lycurgus’ time an acknowledged part of the performance tradition. In turn Lycurgus’ decrees can be understood as a form of public commemoration and cultural restoration for the tragedians rather than as an editorial condemnation of actors. Indeed for Bollack, who does not accept Wyttenbach’s textual addition of παρ᾽, the phrase οὐκ ἐξεῖναι γὰρ αὐτὰς ὑποκρίνεσθαι means that ‘it was not possible to act’ texts of the tragedians prepared only for reading, and that the polis secretary’s job was to help actors prepare theatrical scripts by collating both types of texts.

Bollack’s view of the situation presupposed by the Lycurgan decree is analogous to what I have been arguing in the context of the performance tradition of rhapsodes. Fixed texts of Homer had always provided the backdrop to innovations and extemporaneous flourishes produced in live performance to win over audiences, and it is plausible that such theatrical changes were recorded in later Homeric texts. [48] The late fourth century is, after all, the beginning of the same period in which the Alexandrian rhapsodes and homêristai were experimenting with Homer in hitherto unaccustomed ways, and the record of those experiments survives in the eccentric papyri.