Although the Zenon archive comprises the largest surviving group of documentary papyri from the early Ptolemaic period, the literary and paraliterary papyri known to be connected with the archive are few.[2] Such an absence may dismay those hoping that the intellectual ferment of the third century BCE[3] would reflect itself more directly in one of our best documented figures from that period. Yet among the scanty fragments in the archive, there is one complete poetic work, P.Cair.Zen. IV 59532,[4] a set of two alternative epitaphs for a dog named Tauron.

    This paper will reexamine these epitaphs regarding both their physical placement on the papyrus and the content of the poems themselves, and will seek as far as is possible, given the limited information available, to place them within their poetic and social context. We lack information about the ultimate performance context of these poems – whether they were placed on a stele in public view, recited at a festival, or enjoyed at a symposium. Yet we will see that the papyrus – in its physical appearance and in the poetic form of the epitaphs themselves – connects with outside poetic and cultural structures to convey an ideological message about the two named individuals in the poems – the dog Tauron and his master Zenon. It successfully negotiates the pitfalls of praising an animal to give Zenon ultimately a safe space for his own personal glory.

    P.Cair.Zen. IV 59532:

    Ἰνδὸν ὅδ᾿ ἀπύει τύμβος Ταύρωνα θανόντα
    κεῖσθαι, ὁ δὲ κτείνας πρόσθεν ἐπεῖδε Ἀίδαν·
    θὴρ ἅπερ ἄντα δρακεῖν, συὸς ἦ ῥ᾿ ἀπὸ τᾶς Καλυδῶνος
    4 λείψανον, εὐκάρποις ἐμ πεδίοις τρέφετο
    Ἀρσινόας ἀτίνακτον, ἀπ᾿ αὐχένος ἀθρόα φρίσσων
    [λ]ό̣χμαις καὶ γε̣[ν]ύων ἀφρὸν ἀμ̣εργόμενος·
    σὺν δὲ πεσὼν σκύλακος τόλμαι στή⟦θ⟧‵θ′η μὲν ἑτοίμως
    8 ἠλόκισε, οὐ μέλλων δ᾿ αὐχέν‵α′ ἔθηκε ‵ἐ′πὶ γᾶν,
    [δ]ρ̣α̣ξ̣ά̣μενος γὰρ ὁμοῦ λοφιᾶι μεγάλοιο τένοντος
    [ο]ὐ πρὶ̣ν̣ ἔμυσεν ὀδόντα ἔστε ὑπέθηκε Ἀίδαι.
    [— ˘˘ | ―] Ζ̣̣ή̣[νω]να̣ πονῶ̣ν̣ ἀδίδακτα κυναγόν,
    12 καὶ κατὰ γᾶς τύμβωι τὰν χάριν ἠργάσατο.
    σ⟦κ⟧‵κ′ύλαξ ὁ τύμβωι τῶιδ᾿ ὕπ᾿ ἐκτερισμένος
    Ταύρων, ἐπ᾿ αὐθένταισιν οὐκ ἀμήχανος·
    16 κάπρωι γὰρ ὡς συνῆλθεν ἀντίαν ἔριν,
    ὁ μέν τις ὡς ἄπλατος οἰδήσας γένυν
    στῆθος κατηλόκιζε⟦ν⟧ λευκαίνων ἀφρῶι·
    ὁ δ᾿ ἀμφὶ νώτωι δισσὸν ἐμβαλὼν ἴχνος
    20 ἐδράξατο φρίσσοντος ἐκ στέρνων μέσων
    καὶ γᾶι συνεσπείρασεν· Ἀίδαι δὲ δοὺς
    τὸν αὐτόχειρα ἔθν⟦η⟧‵α′ισκεν, Ἰνδὸν ὡς νόμος.
    σώιζων δὲ τὸν κυναγὸν ὧι παρείπετο
    24 Ζήνωνα ἐλαφρᾶι τᾶιδ᾿ ὑπεστάλη κόνει.
    (m. 2)[τ]ῶι παρ᾿ Ἀ- (vacat) Ζήνωνι

    [A] This tomb tells that Tauron the Indian lies dead, but his murderer saw Hades sooner: the very one which was a beast to behold at close quarters, in fact an immovable remnant from the Caledonian boar (4) was reared in its lair in the fruitful plains of Arsinoe, in the underbrush bristling thickly from his neck, and pressing foam from his jaws. But dashing his strength willingly against the courage of the young dog, the one plowed a furrow, (8) but the other without delay laid (the boar's) neck upon the ground, for when he had grasped a great tendon along with (the boar's) bristles, he did not loosen his teeth until he laid him low in Hades. [He (Tauron) saved (?)] without any schooling Zenon the hunter by undertaking ponos[5] and wrought/fulfilled the charis with his tomb below the earth.


    [B] A young dog, Tauron, who lies buried honorably beneath this tomb, is not without resource against murderers: (16) for, when he met in battle a boar in opposed strife, the one, like someone unapproachable, swelled his jaw and plowed down (the other's) breast, growing white with foam; the other, placing two footholds around the (boar's) back, (20) grasped the bristling one from the middle of his back and folded him to the earth: after he had given his murderer to Hades, he died, as is established law (or: custom) for an Indian (hound). He, saving Zenon, the hunter alongside whom he was following, (24) is sheltered with this light dust.

    [2nd hand]
    [To t]he agent of A-            Zenon
    1 ὅδ᾿ ἀπύει Wilamowitz: ὁδηγεύει Edgar 1920
    2 πρόσθεν: πρόσθ̣ε̣ν Hengstl 1978, Lloyd-Jones and Parsons
    3 ἦ Edgar 1920, 1931: ἢ Wilamowitz
    4 λείψανον, εὐκάρποις ἐμ πεδίοις: λεί̣ψ̣ανο̣ν̣, ε̣ὐ̣κάρπο̣ι̣ς̣ ἐ̣μ π̣ε̣δίοις Lloyd-Jones and Parsons
    6 λ]ό̣χμαις Edgar 1931b: λ]ό̣χ̣μ̣α̣ι̣ς̣ Lloyd-Jones and Parsons: λ]ό̣χμη̣ι̣ Edgar 1920: λ]ό̣χμηι Hengstl 1978
    καὶ γε̣[ν]ύων: κ̣α̣ὶ̣ γ̣ε̣[ν]ύ̣ων Lloyd-Jones and Parsons
    ἀμ̣εργόμενος Edgar 1931b: ἀμ̣ε̣ργόμενος Edgar 1920: ἀμ̣ε̣ρ̣γόμενος Lloyd-Jones and Parsons:
    ἀμελγόμενος Cazzaniga
    7 στή⟦θ⟧‵θ′η Edgar 1931b: στή[ ]η Edgar 1920
    8 αὐχέν‵α′ ... ‵ἐ′πὶ: αὐχέν‵α′ ἔθηκε ἐπὶ Lloyd-Jones and Parsons (a printing error)
    9 δ]ρ̣α̣ξ̣ά̣μενος: δρα]ξ̣ά̣μενος Edgar 1920, 1931: δρα]ξ̣ά̣μ̣ενος Lloyd-Jones and Parsons
    10 ο]ὐ πρὶ̣ν̣ ἔμυσεν: ο]ὐ πρ[ὶ]ν̣ ἔμυσεν Edgar 1931: ο]ὐ πρ[ὶ]ν ἔμυσεν Hengstl 1978: ο]ὐ̣κ ἀ̣ν̣έμυσεν
    Edgar 1920: ο]ὐ̣κ ἐ̣ν̣έμυσεν Wilcken: ο]ὐ̣κ ἀ̣π̣έλ̣υσεν Wilamowitz: ο]ὐ πρὶν ἔλυσεν Page
    11 — ˘˘ | —] Ζ̣̣ή̣[νω]να̣ πονῶ̣ν̣ or: — ˘˘ | —] Ζ̣̣ή̣[νω]να̣ πόνω̣ν:̣ . . . . . . .] Ζ̣̣ή̣[νω]να̣ π̣ό̣ν[ων]
    Lloyd-Jones and Parsons: καὶ σῶισε] Ζ̣̣[ήνω]να̣ πο̣ν̣[ῶν] Hengstl 1978: σώιζει δὲ] Ζ[̣ήνω]να πόνων Page
    12 ἠργάσατο Edgar 1920, 1931: εἰργάσατο Cazzaniga
    Between lines 12–13: paragraphos mark
    13 ἄλλο in ekthesis
    22 ἔθν⟦η⟧‵α′ισκεν Edgar 1920: ἔδν⟦η⟧‵α′ισκεν Hengstl 1978 (a printing error) Ἰνδὸν (ς corrected to ν)
    Edgar 1931b

    25–26 τ]ῶι παρ᾿ Ἀ[π]ολλωνίου: τῶι παρ᾿ Ἀπολλωνίου Edgar 1920, 1931b (The first letter of each line is hidden behind the mounting)

    Zenon's personal connection with this set of poems has excited a moderate amount of interest among modern scholars, but much of the discussion on these poems has dwelt on what has become a set of binary positions. These poems are said to be either a sort of exemplar to guide Zenon in commissioning a permanent monument in stone, or, conversely, they are a literary exercise unanchored to actual events. The lack of any information outside of these poems makes arguing either side futile, since the side in favor of an inscription can only offer arguments from plausibility, while those opposed support their thesis with two equally unverifiable positions. The first strand depends on arguments ex silentio, such as the observation that no dog epitaph survives in the archaeological record before the first century CE, but such absence of evidence proves nothing by itself, especially when the archaeological remains of Philadelphia have been obliterated in modern times.[6] The other major strand involves finding details in the poems at odds with the contemporary "reality" of the third century Fayum, such as Gortemann's observation as to the implausibility of a boar being mentioned as roaming the Arsinoite nome,[7] or Purola's insistence that finding "stereotypical" elements within these poems points to their exclusively literary purpose.[8]

    That is not to say that we must reject the literary elements that others have found, which we will cover later. Yet we must keep in mind that there is ample conceptual space between an honest reportage of events and pure literary invention, and that the truth of Tauron's actions or even his existence is ultimately immaterial to the inscriptional or literary character of these poems. Both media, in fact, show considerable cross-pollination during the Hellenistic period, as both Wilhelm and Robert remarked long ago.[9] Perhaps the greatest drawback of this obsession with the truth-value of these poems is that it takes the measure of historicity in light of a distinctively modern obsession with the honest reportage of events. Consequently, it misses crucial points of contact between those events and the ideological representation of them. To give a short illustration from the area of funerary commemoration, Fantuzzi and Hunter remark that an epitaph dedicated to a Menophila is "no less literary than the epigrams of the Palatine Anthology" – yet this epitaph was not found in a collection of poetry, but instead in the city of Sardis, on a grave stele dated to the late second century or early first century.[10] Conversely, the Atthidographer Philochoros' lost collection of Attic epigrams in the late fourth or early third century shows that existing epitaphs could enjoy a literary circulation and serve as models of epigraphic practice for poets and readers.

    Two key pieces of information connect this papyrus to Zenon. It is linked with the Zenon archive first by reason of its provenance and secondly by another key piece of information (Fig. 1; ll. 25–26): the address that exists on the verso ([τ]ῶι παρ’ Ἀ/[π]ολλωνίου || Ζήνωνι "To the agent of Apollonius, Zenon"). This address suggests that we have the original set of poems sent to Zenon. The papyrus shows no sign of reuse, and the similarity of the hands between the recto and verso of the papyrus makes it unlikely that these poems were copied onto a blank papyrus that was addressed to Zenon. If, however, we believe this is a duplicate copy deposited in Zenon's archives, we are hard-pressed to answer why such an exacting copy on expensive papyrus was needed, and why, if this is a record kept for reference, more information was not appended to it. The papyrus, as De Luca observes, is of fine quality, and consists of three sheets with a total dimension of 33.7 x 39.1 cm joined by two kollēseis.[11] Only the middle sheet is uncut, whose width of approximately 21.5 cm approximates that which Johnson 1993 has calculated for the "hieratic" grade of papyrus, if Pliny the Elder's account is to be trusted.[12] If we examine the hand of the address on the verso, however, we can make one more connection (Fig. 1). Edgar in his edition of P.Mich. I 77 (an Apollonios[13] to Zenon) briefly entertains the possibility that the author of this letter is identical with the poet of P.Cair.Zen. IV 59532, but goes on to reject the identification on the grounds that the hands are similar, but not the same. It appears that Edgar's rejection was based on a comparison of the hand on the recto (which was the only published photo at that time) against the documentary text, but the address on the verso in fact makes a much better match (Figs. 1–2). There is some variation, however, between the forms of nu and eta between 59532v and 77r. Yet we find that nu letter forms and omega-nu ligatures on 59532r (l. 24) and elsewhere on 77r (l. 5) correspond quite closely, while an eta in line 6 of 77r exhibits a much less angled crossbar that connects to the midpoint of the right hasta rather than the top as elsewhere – somewhat more like the etas in 59532. Though there is a small possibility that the hands of the recto and the verso of 59532 are different, a much more plausible explanation in light of their general similarity is that the poems on the recto were written more slowly and carefully than usual, and that the writer reverted to a faster documentary hand when it came time to write the address. The valediction of P.Mich. I 77.16 also shows an apparent second hand. While it may be tempting to posit that we have an instance here of an amanuensis composing the bulk of both letters, letters with multiple hands are exceedingly rare in the Zenon papyri, in direct contrast to the common later documentary practice of having the author append a subscription to a letter written by a scribe.[14] Though these variations in hand are commonly understood to signify a change in the person writing, all these variations are in short valedictions, which arguably are such a frequent element of letters that a writer could develop a stereotypical way of writing them independent from his regular hand, as seems to be the case in P.Lond. VII 2033 (Epharmostos to Zenon).[15] Deliberate variation in writing style is visible elsewhere in the Zenon archive in P.Cair.Zen. IV 59535, a practice copy of lines of poetry and oratory attributed to Epharmostos, the brother of Zenon.[16] In other words, we have different hands here but the same author. Since our papyrus has no identifying information about the sender either on the verso or within the poems themselves, it follows that this letter was expected by its recipient, Zenon, and that, much like the arrangement implied in Apollonios' other letter, P.Mich. I.77, some sort of financial arrangement or favor was being done between Apollonios and Zenon. Since this piece lacks a docket describing its purpose, 59532 could appear to have crept into the archive accidentally, perhaps even for sentimental reasons. Yet an argument for sentimentality should be a last resort rather than the first explanation, since it presumes emotional attachment instead of seeking other reasons for the papyrus' inclusion in the archive – and we may add, its ultimate exclusion in being discarded with the other pieces that have come down to us. The importance of understanding this piece's poetic place and the uses it may have had is especially acute if this piece is connected with an individual with whom Zenon conducted correspondence.

    Fig. 1:: P.Cair Zen. IV 59532v. Image of address on the recto (ll. 25–26 of text).
    Fig. 1:
    P.Cair Zen. IV 59532v. Image of address on the recto (ll. 25–26 of text).
    Fig. 2:: P.Mich.Zen. I 77r. Image of lines 1–4 and 14 on the recto; letter forms similar to the address on P.Cair.Zen. IV 59532v have been underlined with color-coded bars in this and the previous figure.
    Fig. 2:
    P.Mich.Zen. I 77r. Image of lines 1–4 and 14 on the recto; letter forms similar to the address on P.Cair.Zen. IV 59532v have been underlined with color-coded bars in this and the previous figure.

    The papyrus itself is of very fine quality, and of dimensions that would more readily suggest a bookroll than a personal letter. The lack of any information on the recto external to the texts other than the ἄλλο in line 13 very likely mimics the appearance of a contemporary collection of poetry, for Fantuzzi and Hunter note that the practice of identifying poems individually was far from standard in the Hellenistic period, to judge from the Milan papyrus attributed to Posidippus (P.Mil.Vogl. VIII 309) and the "obvious improvisation" of headings in poems attributable to earlier Hellenistic collections preserved in the Greek Anthology.[17] Most importantly, though, an accident of preservation enhances our understanding of the text – namely, the verso shows signs of offsetting, the bleeding of ink from the recto to the verso.[18] From the information that the offset provides, we can ascertain a number of facts about the geometry of the rolling of the papyrus at the time of the offset. I have provided a summary of the steps and the relevant equations in Figs. 3–10. First, the papyrus was fairly evenly rolled at the time of the offset, with the rolling angled at only 1.72 degrees (Figs. 3–4). Correcting for the angling and then reading the overlaid texts together has allowed for most of Edgar's 1930 readings to be confirmed against later conjectures, though there is not enough space to discuss each of them in detail. Next, the inner diameter of the roll, di, and a diameter from about the middle of the roll, dm, can be found (Figs. 6–8). The rather large value for the inner diameter of the roll, 1.4 cm, shows us that the papyrus had been unrolled and presumably read by the time of the offset. From two diameters on the roll, we can roughly calculate (Fig. 9) the thickness of the papyrus, which works out to .214 mm, half of the typical papyrus thickness that Johnson specifies for the Roman period.[19] A few Roman-period literary papyri from Tebtunis in Berkeley's collection show a thickness comparable to the one here, so we must accept it as plausible.[20]

    Fig. 3:: P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the left margin of the text on the recto. Text of the recto shown in black; offset text of the verso shown in white.
    Fig. 3:
    P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the left margin of the text on the recto. Text of the recto shown in black; offset text of the verso shown in white.
    α = tan-1(h/w) = tan-1(3.96 mm/131.58 mm) = 1.72º
    Fig. 4:
    The angle of folding between the recto and the verso (angle α in Fig. 1) can be calculated by taking the inverse tangent of the proportion of the height of the divergence between the lines (h) to the width over which the divergence occurs (w):
    Fig. 5:: P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the left margin of the text and rotated 1.72º. Note that the white offset text of the verso corresponds vertically with the black text of the recto. The horizontal divergence of the texts is the result of the compression of the offset text on the rolled/folded verso during the time at which the offset took place.
    Fig. 5:
    P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the left margin of the text and rotated 1.72º. Note that the white offset text of the verso corresponds vertically with the black text of the recto. The horizontal divergence of the texts is the result of the compression of the offset text on the rolled/folded verso during the time at which the offset took place.
    Fig. 6:: P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the right edge of the recto. ki is the inner circumference of the roll at the time of offsetting; km is the largest outer circumference within the roll that can be ascertained from the offset; W1 is the length of the roll at the point of that outer circumference; W2 is the total width of the papyrus (= 33.7 cm).
    Fig. 6:
    P.Cair.Zen. IV 59532. Image of lines 1–3 on the recto with a superimposed false color mirror image of the verso aligned to the right edge of the recto. ki is the inner circumference of the roll at the time of offsetting; km is the largest outer circumference within the roll that can be ascertained from the offset; W1 is the length of the roll at the point of that outer circumference; W2 is the total width of the papyrus (= 33.7 cm).
    di = ki/π = 42.80 mm/π = 13.62 mm
    Fig. 7:
    The inner diameter of the roll (di) at the time of the offset is calculated by finding the circumference at the inner part of the roll (ki), which can be taken by measuring the distance between the right margin of the offset text on the mirrored verso and the right margin of the text on the recto when the verso is superimposed and aligned to the right edge of the recto (see previous figure).
    dm = km/π = 48.67/π = 15.49 mm
    Fig. 8:
    The diameter of the roll at a middle point within the roll (dm) can be calculated by finding the circumference of the roll at the left margin of the text (km), which can be taken by measuring the distance between the left margin of the offset text on the mirrored verso and the left margin of the text on the recto when the verso is superimposed and aligned to the right edge of the recto (see Fig. 6).
    t = (π/W1) · ((dm/2)2-(di/2)2) = (π/200.69 mm) · (15.49 mm/2)2-(13.62 mm/2)2 = 0.214 mm
    Fig. 9:
    Calculation of approximate thickness (t)[21] by using the data we have for the inner diameter (di), outside diameter (dm), and length (W1) and plugging the values into the equation for calculating the thickness of an ideal roll (see Johnson 2004,149–150 n. 70 for an analogous use of this equation).
    Fig. 10:: Using the equation in the previous figure to solve for the diameter given the thickness, thus yielding the minimum diameter of the roll (Dmin; i.e., the minimum diameter the roll can have if it was rolled as tightly as possible during the offset.
    Fig. 10:
    Using the equation in the previous figure to solve for the diameter given the thickness, thus yielding the minimum diameter of the roll (Dmin; i.e., the minimum diameter the roll can have if it was rolled as tightly as possible during the offset.

    The folds in the surviving text and the angle of the offset show that the papyrus was rolled from the right to left, much like a bookroll, rather than in the normal way of rolling a private letter, which was from top to bottom. That resemblance to a bookroll is further heightened by a left-hand margin of approximately 11.1 cm, almost a third of the total length of the papyrus. This wide margin, created by the sheet joined a little left of the left margin of the text, forms a sort of prōtokollon that imitates a book roll's protective outer sheet of blank papyrus. Though the hand is primarily documentary, and the vertical margins are not overly generous, the lines of text have a gentle and more-or-less constant rise across the fibers of the papyrus. The symmetrical margins on the bottom and the top of the page are achieved by slightly increasing the size of the letters at the bottom of the column, and suggest that, for an experienced hand such as the one we have here, the text was composed on the sheet of papyrus rather than copied from another.[22] The correction in line 22 to Doric dialect also hints at a last-minute change to harmonize with the Doric dialectical features of the rest of the poems, possibly in reference to the Carian origin of Zenon. The erasure of the nu in line 18 is not necessarily an error in copying as De Luca suggests, but rather could be the result of a substitution during composition of the metrically superior λευκαίνων ἀφρῶι for ἀφρῶι λευκαίνων, which is the order in which these two words appear together in Nic. Alex. 170 (a later text, but one that could be drawing on either a fixed expression or an earlier work).[23] Another set of corrections, as De Luca notes, are to improve the legibility of the text, and thus point to the importance of this text being composed to be read rather than as a school exercise: σ⟦κ⟧‵κ′ύλαξ in line 14 and στή⟦θ⟧‵θ′η in line 7 have had letters added above the line when dribbling ink has interfered with their legibility. Finally, the corrections to line 8 have led Lloyd-Jones and Parsons to suggest that they are the sign of an inexperienced poet, while Saija suggests they are a representation of synaloiphe, though the addition of the epsilon above ἐπί makes that explanation problematic.[24] A better explanation, perhaps, is that they are a correction to give the pieces a special type of scriptio plena in which elision is represented only with non-lexical words (as in ll. 1, 3, 5, 8, 14, 15, 19, and 24), a practice found regularly in prose inscriptions and sometimes in verse ones.[25] This style, along with the setting of the text in one, block-like column, was meant to evoke the appearance of an epitaph even if its final destination was not stone, and to grab its reader like an epitaph would capture the eye of a passerby.

    The corrections to the text may tempt us to think that this poem is somehow a careless, slipshod composition,[26] but we must remember that we have no other autograph poem from this period with which to compare it. For all we know, a clean copy could have been thought to be needed only at the text's "publication," whether that was as inscribed epigram, public or private reading, or circulated text. Further, what may appear to be metrical difficulties in the two poems are more the result of our reliance on the top levels of Alexandrian poetics (especially Callimachus) for our understanding of metrics than any lack of versificatory skill on Apollonios' part. We find masculine caesurae here more frequently than in Callimachus and Theocritus, yet such a preference is also found in Aratus.[27] Perhaps more unsettlingly, scansion of Ἀίδαι in lines 10 and 21 must be different (anapestic in the first instance, cretic in the second). While the cretic scansion of this word is highly unusual, it is attested securely once at Semon. 7.117 (Ἀίδης), whose pattern Apollonios may be following for its scansion in iambics.[28] Most troublingly, we see masculine caesurae after two so-called "second trochees" in lines 3 and 9. The first of these is less of a problem, because ἄντα and δρακεῖν are tied so closely together syntactically that they function like a single word. The situation in line 9 cannot be explained away, however. Yet we only have to expand our metrical perspective a little further among Apollonios' contemporaries to see a similar example in elegiacs, AP VII 50.3 (Archimelus). If we consider that we only have the most popular epigrams of all but the most famous poets of this period, it may be easier to see the poems here as nearer to the mainstream of middlebrow Hellenistic poetry than the product of a clumsy school exercise.[29]

    The structure of both epitaphs broadly follows the conventions of epideictic in constructing the praise of Tauron. In both epitaphs, the first two lines introduce Tauron's achievement in the abstract – namely, that he is able to mete out justice to murderous boars. Then both poems go on to describe that achievement in terms of his qualities (his aretai – implied by τόλμαι in line 7 of the first epitaph, and then by Ἰνδὸν ὡς νόμος in the second) and his concrete actions (his praxeis – his exceeding the boar's strength in lines 8–10 of the distichs and his stopping and defeating the boar in battle in lines 19–22 of the trimeters). Both poems give a kind of genealogy in naming his type/ancestry[30] (Ἰνδόν[31] in ll. 1 and 22), and set up a consolation passage in the final two lines by linking Tauron's heroic deed with those who survive him, namely Zenon as hunter, to whose oikos Tauron belongs. The action against the boar is transferred to an act that saves the life of Zenon, set up in the first poem as an exchange of charis in Zenon's (at least literary) creation of a tomb, and in the second with a variation on the traditional wish that the earth rest lightly on him. Since the detail of Zenon's life being saved is confined to the consolation rather than the description of events, the danger of reading Zenon's needing to be saved as a rebuke is entirely averted. In both poems, Tauron is identified as a young dog (σκύλακος in line 7 and σκύλαξ in line 14), and his youth makes the description of his good qualities and deeds especially apt for exaggeration (auxēsis), in essence following a prescription in the near-contemporary Rhet. [ad Alex.] to exaggerate the achievements of the young, since they provide less material for praise.[32] The last two lines of the couplets construct Tauron's action as ἀδίδακτα, without any schooling, appropriate for a young subject of praise and for an animal who is not in the strict sense taught to save his master, but also appropriate for amplifying Tauron's achievement. Though it may be tempting to read πόνων ἀδίδακτα here, taking πόνων as a relational genitive dependent on ἀδίδακτα,[33] we must remember that ἀδίδακτα is operating as an adverb. ἀδίδακτος as an adjective can take a relational genitive, but adverbial compounds of διδακτός do not take dependent genitives.[34] That is not to say that we should ignore the paradox of the rhetorical placement of these two words, particularly because the only difference between these readings is whether the contonation or rise-and-fall of the pitch accent begins on the omicron or the omega (in either case, the pitch accent completes its fall on the omega). This linkage instead magnifies the achievement of Tauron's actions against what should be his nature, for animals were proverbially ἀδίδακτα.[35] Moreover, the opposing boar is practically made into a mythological monster, as is overtly clear in the mythological comparison with the Caledonian boar in lines 3–4 and multiple times by words such as ἀτίνακτον ("immovable," line 5) and ἄπλατος ("unapproachable," line 17). This exaggeration of the boar as a threat to civilization is particularly notable when we remember that the poet has essentially turned the subject of the poem, Tauron's accidental death while hunting – while actively seeking to kill game such as boar – into a heroic, defensive move.[36] Both poems thus have a formulaic structure: two lines of Tauron's achievement in the abstract, followed by 4 or 3 lines exaggerating Tauron's opponent (this difference serves to let the couplets have the required even number of lines), 4 lines about the event that return to the idea that Tauron gave a murderer its due, and 2 lines of consolation framing the act as one that saves his master.[37] In addition to this episodic structure,[38] both poems follow the general pattern that Menander Rhetor prescribes for the composition of monody:[39] the first two lines in both poems mention Tauron's present deceased state, the middle portions narrate the achievements of the dog in the past, while the final two lines of each look to the future of Zenon's consolation and the abiding presence of poem and "tomb."

    Theme Distichs Iambs
    General statement of Tauron's achievement (cf. Anax. Rhet. 35.11–12 [1441a]; Men. Rh. II 420.10–12 [epitaphios]); present time lines 1–2 lines 14–15
    Description of the boar with auxēsis to increase praise of the young deceased Tauron (cf. Anax. Rhet. 3.1 [1425b], 35.12 [1441a]; Men. Rh. II 435.1–7 [monodia]); past time lines 3–6 lines 16–18
    Narration of the event; (moral) statement of the boar receiving his due (cf. Anax. Rhet. 3.2,6 [1426a], 35.15 [1441a]; Men. Rh. II 420.21–27 [epitaphios]); past time lines 7–10 lines 19–22
    Consolation to Zenon (cf. Men. Rh. II 419.5–6, 421.14–16, 422.1–2 [epitaphios]); future time lines 11–12 lines 23–24

    In short, we can see that the structure of these works reveals a poet highly conversant with the genre of epideictic, and thus with a solid oratorical training, perhaps no surprise for a man who quotes literature freely in his personal correspondence and has dealings with a teacher.[40] Yet we must deal with an important aspect of these poems of praise – that they are expressed in the most fulsome terms not to a young man or even a child, but to a dog.

    The praise of animals has a long tradition in Greek literature, from the Argos episode in the Odyssey, to Pindar's and Bacchylides' praise of the horse Pherenikos within their epinicia to Hieron I, to later school exercises in which students were asked to compose praise of animals.[41] Yet we cannot date any animal epitaph securely before the late fourth or early third century, if one recognizes that Page's dating of Pseudo-Anacreon's horse epitaph[42] is entirely conjectural. A brief passage in Aristotle's rhetoric (2.24.2 [1401a]) reveals a reason for this reticence: there is a distinct danger in praising animals that one can go too far and make the praise seem insincere, ironic, or paradoxical – an effect that one could argue is a feature of the later school exercises.[43]

    In fact, this very danger of taking praise of animals as ironic is realized in Cazzaniga's reading of 59532. Starting from the somewhat bizarre assertion that ἀπύει in line 1 can be nothing other than an allusion to the Knights of Aristophanes, despite his being virtually unread during this period, Cazzaniga characterizes the poems as parodistic and full of humor. Yet if we take ἀπύω in its rightful place as a high-register word for communicating,[44] Cazzaniga's reading depends on taking the exaggeration of this passage as ironic – but we have seen that this exaggeration is a feature of epideictic. Instead, these epitaphs work with full power as epideictic because of Tauron's status as a dog, a status that was invested with great and often contradictory meaning in Greek culture.

    Cristiana Franco demonstrates that the dog had a unique place among animals in being a part of human society (though on the lowest rank) and thus (however unrealistically) being expected to abide by human mores.[45] The upshot of this conception was that the praiseworthy qualities of dogs (daring, fidelity, etc.) when they were subordinated to the master became their worst qualities (recklessness, faithlessness, etc.) once they left patriarchal and hierarchical control. The obedient dog could become the ideal subordinate, in that it reflected the wishes of its master, and, as such, a status symbol.[46] The dog could signify both patriarchal power as his master's agent or the lowest level of subordination as his master's object. We will see Tauron's potential fulfilling of these two functions if we examine the recto and the verso of the papyrus separately according to the two different social spaces in which each side traveled and different degree of control Zenon could be presumed to exercise over others’ access to the outside and inside or a sealed letter. The recto was destined to be opened in Zenon's household, where Zenon had notional control over its dissemination and its performance in praise of one departed member of his household, and, ultimately, of the household itself.[47] The poems construct social relationships on an equal level and on a heroic pattern – and even the antagonistic relationships are figured in this manner. The boar makes his appearance in the couplets through a mythological comparison, and is introduced in the second poem within an exemplum (introduced by γάρ in l. 16) that illustrates the gnomic assertion that Tauron is ἀµήχανος in line 15. The word ἀµήχανος itself frequently has the meaning "irresistible" in Homer and other archaic poetry, separate from its meaning of "helpless" in later Greek.[48] The use of ἄπλατος (unapproachable), a synonym of the Homeric ἀµήχανος, two lines later offers the potential of it being a gloss to Homeric definition until we realize later in the line that it is being applied to the boar. This semantic effect parallels the balanced sets of lines (17–18 against 19–20) as well as the syntactical balance in the μὲν ... δέ construction, with the result that both the boar and Tauron momentarily occupy the space of the unassailable. Yet the solution to the inherent grammatical paradox lies in the other definition of ἀµήχανος: Tauron is not helpless against the boar as potential murderer of Zenon (cf. ἐπ᾿ αὐθένταισιν in l. 15), but only at the cost of the dog's life. The narrative thus activates both definitions: Tauron proves to be οὐκ ἀµήχανος not only because he protects Zenon effectively, but also because that protection is not irresistible. It is precisely balanced against the boar’s own force. That horizontal equilibrium continues in the consolatory conclusions to both poems, for Tauron accompanies Zenon the hunter as an equal companion (l. 23) and his relationship with Zenon is represented in lines 11–12 as the equal exchange (the charis) of the tomb for his ponos (his heroic labor of defeating the boar and poetic labor of creating material for the poem). Zenon is never referred to as master – instead, the noun κυναγός (hunter or literally, leader of dogs) is used for him. Zenon's house is represented as a realm of equality and aristocratic exchange.[49]

    On the other hand, the verso places Zenon squarely in the employ of Apollonios, which, though it is not a true subordination, represents a hierarchy. Reading the poems from that side of the papyrus, we see Tauron the dog and Zenon his equal take on another set of qualities – those of the dog being the perfect subordinate. The dog's unique place in Greek society allows Zenon to participate in a glorification of his dog and his house without running the risks of asserting himself too boldly as being a member of Apollonios' elite rather than his own local, Philadelphian one. Zenon's attempt here to create a space for safe self-glorification is thus analogous to what Richard Neer has observed the vase-painters of the Athenian Pioneer Group doing when they paint themselves in gymnasia and symposia with the trappings of the elite.[50] The space that a dog offers for self-glorification is perhaps what lies behind one of the criticisms of the man with petty ambition (μικροφιλοτιμία) in Theophrastus' Characters.[51] The viewpoint of this critique is unabashedly elite, in that it treats the self-glorification of a parvenu like Zenon as failing miserably because the dignity of its inscription (which takes the form of the epitaph of a metic[52]) is not modulated adequately to its being applied to a dog. Perhaps all that was lacking for making it a powerful yet safe act of self-glorification was a poet like Apollonios.


      1. My thanks to Professors Todd Hickey and Leslie Kurke for reading previous versions of this paper. I also would like to thank Professor Anthony Bulloch for offering me feedback on my interpretation of these poems' content and metrics. Any shortcomings that remain are of course my own.return to text

      2. The relative lack of literary papyri connected with the Zenon archive could be the result of their being separated or kept separate from Zenon's "library" in antiquity (Clarysse 1983, 53), or possibly because after their undocumented excavation, the more valuable literary texts were taken from the main body of the archive for separate sale and the documentary texts were selected out by Edgar, Nahman, and others acquiring for collections at Cairo and elsewhere. See Hickey 2009, 71 for a similar selection process by language (Demotic, Egyptian, Greek) occurring in collections of papyri from Tebtunis. P.Zen.Pestm. 14 and 15 are two such texts sold as individual pieces but later connected to the archive; the former was linked to the archive by the hand of the account on its verso, the latter by the circumstances of its acquisition. P.Zen.Pestm. 16, whose acquisition record identifies it as coming from the Zenon archive, is probably another such text, though there is some uncertainty whether the form of the letter xi in the document on the verso might date the document too late (P.Zen.Pestm. 16 descr.). Yet since it is likely that the text on the recto is prior to the verso, and since the document may have been reused at any point before being discarded, the existence of a later text on the verso may be less of a problem, though no evidence exists for an unassailable link. All the other literary or paraliterary papyri sold as part of the archive (P.Cair.Zen. IV 59532–59535) are not written in easily identifiable literary hands. Further, the fact that the archive itself comes from undocumented excavations in Philadelphia is only hearsay, as Viereck 1928, 5 is our earliest source for their provenance (presumably on the authority of the dealers, who may have had an interest in concealing their true source). Whether other undocumented literary papyri from the third century Fayyum could be linked with the archive remains a matter of supposition and possibly further investigation.return to text

      3. Unless otherwise noted, all subsequent dates are BCE.return to text

      4. For images of the recto and the verso of this papyrus and many of the other papyri discussed in this paper, refer to the Photographic Archive of Papyri in the Cairo Museum at: <>(accessed 1 May 2008).return to text

      5. Or (less likely): "without any schooling in ponos," if one prefers the reading πόνων. See discussion below.return to text

      6. Further, in Day's catalogue of archaeologically attested Greek dog burials (Day 1984, 22–25), a fourth cent. and a late Hellenistic burial are found, though neither of these survive with inscriptions. For a recently published (though later) dog burial found together with an inscription, see Merkelbach and Stauber 1998, no. 17/15/01.return to text

      7. Gortemann 1957, 116–118.return to text

      8. Purola 1994, 62.return to text

      9. E.g., Wilhelm 1900, 94; Reichel and Wilhelm 1901, 74–75, and Robert 1960, 588 n. 4 note third cent. inscriptions with letter forms that mimic literary texts. For a recently published Greek inscription from Afghanistan, possibly from the late Hellenistic period, that has a blank left margin in imitation of a book roll, see Bernard 2004, 238 n. 13.return to text

      10. Fantuzzi and Hunter 2004, 337–338 on Merkelbach and Stauber 1998, no. 04/02/11.return to text

      11. De Luca 2002, 260; Edgar 1931a records the dimensions as 34 x 39.5 cm, though it is impossible to tell whether the differences between these two figures is the result of De Luca's more exact measurement or shrinkage of the papyrus in the seventy years between publications.return to text

      12. magna in latitudine earum differentia: xiii digitorum optimis, duo detrahuntur hieraticae, Fanniana denos habet, et uno minus amphitheatritica, pauciores saitica, nec malleo sufficit; nam emporiticae brevitas sex digitos non excedit. (Plin. HN 13.78) (Trans. Bostock and Riley, slightly adapted, with Johnson 1993's estimates in brackets) There is a great difference in their breadth. That of best quality is thirteen fingers wide [24.1 cm], while the Hieratica is two fingers less [20.4 cm]. The Fanniana has (a measurement of) ten fingers [18.5 cm], and the Amphitheatrica one less [16.7 cm]. The Saitic is of smaller breadth and is not as wide as the mallet (with which it is beaten); and the narrowness of the Emporetica does not go beyond six fingers (in breadth) [≤11.1 cm].return to text

      13. But not the dioikētēs, since the hand is entirely different. This Apollonios, however, is an excellent candidate for the author of the poems here, as in P.Mich I 77 he has dealings with at least one didaskalos (5–6) and appears to make a literary quotation (12ff). return to text

      14. Evans 2005, 196; Evans 2004 and 2005 list a total of 5 letters with variations in hand: P.Lond. VII 2019 and 2033, P.Zen.Pestm. 51, P.Cair.Zen. II 59283 and 59284, to which we should add the ones here and P.Cair.Zen. V 59808 (the valediction and the date are in different hands, according to Edgar's description).return to text

      15. As first observed by Skeat in P.Lond. VII 2033.descr.return to text

      16. See Edgar 1931a, P.Cair.Zen. IV 59535 descr.return to text

      17. Fantuzzi and Hunter 2004, 302.return to text

      18. Offsetting happened at some point when the papyrus was rolled up. Though Lloyd-Jones and Parsons, 977 descr. claim that the offset is the result of the modern unrolling of the papyrus, the fact that the docket of P.Lond. VII 1941 is offset on P.Cair.Zen. I 59061 (according to Skeat in P.Lond. VII 13–15n. both papyri appear to have been acquired in different lots) suggests that the offsetting process happened sometime before the papyri were excavated.return to text

      19. Johnson 2004, 149–150 n. 70. Johnson in arriving at this figure actually specifies it as the "interval," or distance between successive layers of papyrus in a bookroll, which may or may not correspond exactly to the thickness of the papyrus. In his figure accompanying the note, however, he labels it as equivalent to the roll's "height" (my thickness). In any but the most heavily used rolls, the even compression of the inner surface of the roll (the recto) would tend to keep the layers tight against one another and close to the actual thickness of the papyrus itself.return to text

      20. I was unable to locate a Ptolemaic-period papyrus among the Tebtunis collection at Berkeley that was available for measurement (i.e., not mounted in glass), but the following measurements reveal that .214 mm is a plausible thickness, particularly for the finer grades of Ptolemaic-period papyri: P.Tebt.Suppl. 1017: .33 mm; P.Tebt.Suppl. 1018: .21 mm; P.Tebt.Suppl. 1032–1033: .21 mm; P.Tebt.Suppl. 1206: .48 mm; P.Tebt.Suppl. 1558: .29 mm; P.Tebt.Suppl. 1715: .27 mm (all likely second – early fourth century CE).return to text

      21. It would represent the actual thickness of the papyrus if the roll at the time of the offset were rolled as tightly as possible between di and dm. Unfortunately, the tightness of this rolling cannot be reconstructed without information on the actual thickness of the papyrus, which is impossible to calculate without removing the papyrus from its frame.return to text

      22. Thus it follows Maas' Law (see Johnson 2004, 91–93).return to text

      23. De Luca 2002, 261. Note that Suetonius records seeing a set of corrections reflecting the composition process in a manuscript of Nero’s poetry, from which he draws the conclusion that it represents an autograph copy (Nero 52).return to text

      24. Saija 1990, 29–34.return to text

      25. Devine and Stephens 2004, 262–263. Note that while ὅδ᾿ in 1 is a lexical word, the elided epsilon carries no lexical meaning (the word being formed from the lexical demonstrative pronoun ὁ and nonlexical enclitic – δε·: see Schwyzer 1.610–612, esp. 611–612 β).return to text

      26. E.g., in the words of Lloyd-Jones and Parsons, the poet is ineptus, sed non indoctus (SH 977 descr.).return to text

      27. Kidd 1997, 33.return to text

      28. Contra Cazzaniga 1973, 88 n. 30, he is less likely to be following Soph. OC 1680, which suffers from a textual problem.return to text

      29. As Cribiore notes (2001, 240), it is often difficult to differentiate between advanced school exercises and professional products. If we accept that this papyrus is the product of the Apollonios who wrote P.Mich. I 77, it is very likely a professional product, as he appears to have finished his schooling.return to text

      30. Understanding the different types of ancient dogs as "breeds" is extremely anachronistic, as our modern conception of the breed depends on closed breeding lines and studbooks that were not a feature of the ancient landscape, in which geographic origin and ultimately outward appearance were the chief determining factors. See Sundqvist 2006, 1127 for genetic evidence that all dog breeding before the late eighteenth century CE was "almost at random."return to text

      31. Contra Orrieux 1983, 136 with 156 n. 17, the end of line 22 does not refer to the Indian practice of auto-immolation. His reading is dependent upon reading Ἰνδός in the text when it has clearly been corrected to Ἰνδόν, and his citation of the Calanus legend in Arr. Anab. VII 3 and Plut. Alex. 69–70 has little relation to the text here. Diodorus Siculus and Strabo, however, record a story of Alexander and Indian dogs that features an Indian dog keeping a hold on its quarry even when a leg is cut off by a swordsman (Diod. Sic. XVII 92 and Strabo XV 31). Even though those accounts are more appropriate to the sentiment here, they have all the hallmarks of an invented aetiology rather than a true piece of folklore. I will examine in a later study the full significance of the Indian dog in Greek, Persian, Indian, and Chinese sources, but for now it will suffice to say that the animal has a close connection to Persian elite practices and from there enters Greek discourse as an eastern luxury item (attested by Xen. Cyn. 10.1 as one of the dogs suited to hunt boar; this belief may explain why Tauron is paired against a boar here). Callixenus' description of the procession of Ptolemy II (Athenaeus V 32b = FGrH 627 F2) shows them within a roughly contemporary context of Ptolemaic royal display.return to text

      32. Anax. Rhet. 35.12 [1441a] (transmitted among the works of Aristotle)return to text

      33. See footnote 5, above.return to text

      34. Thus it is unlikely that these words are some type of school-themed joke, as Cazzaniga suggests (1973, 87). In fact, Cazzaniga's "translation" ἀμαθῶς (ibid.) does not take a dependent genitive unless it is paired with ἔχω (which is a special case). We should not lose sight of the fact that the social meaning of ponos is often tied closely to the achievements of hunting and athletics (Loraux 1995, 48, describing Xenophon's use of the term; 50 with references in n. 41, especially as opposed to truphē). We thus must be careful not to read its use here automatically as literary-critical.return to text

      35. See Hp. Alim. 39: Φύσιες πάντων [τῶν ζῴων] ἀδίδακτοι (other versions of this proverb make it clear that πάντων is understood to be possessive genitive agreeing with understood animals).return to text

      36. At least according to Xenophon (Cyn. 10.8, 20–21), the Greek practice of boar hunting involved using dogs to draw the boar's attention and the danger away from the hunter, a tactic which often resulted in the death of many dogs.return to text

      37. An analogue in traditional epideictic would be the assertion that men died to save their city, e.g., Thuc. II 44.return to text

      38. Arguably, each of these four sections could illustrate the four cardinal virtues: the first section reflects Tauron's justice (in punishing his own murderer), the second section reflects his courage (his opponent is a monstrously fearsome boar), the third section reflects his temperance (he behaves as a warrior should: for this sense, see Xen. Mem. III 5.21, where σωφρονεῖν is paired with εὐτακτεῖν), and the final section reflects his wisdom (his piety in saving Zenon).return to text

      39. Men. Rh. II 436.16–436.10.return to text

      40. See above, footnote 13.return to text

      41. See Cribiore 2001, 139 n. 36, and 229 n. 39.return to text

      42. Page 1981, 401–402.return to text

      43. Though Aristotle implies that it is impossible for the praise of animals not to be paradoxical, and Polycrates' praise of mice (Aristot. Rhet. 1401b15, Demetr. de Eloc. 120 with Radermacher 1951, 131) is clearly ironic, the acceptance by Anaximenes that the subject of praise can be ἄνθρωπον ἤ τι ἄλλο τοιοῦτον ζῶιον ("a person or some other animal of such a sort," 35.5 [1440b]) signals that, by the late fourth cent. at least, composing praise for animals was not in itself ludicrous. Of the animal epitaphs preserved in book VII of the Palatine Anthology, the parodying and ironic poems date from the late Hellenistic period (Herrlinger 1929, 94 further notes that these Hellenistic-period poems come from authors who write within the Greek cultural sphere yet ultimately hail from outside of it).return to text

      44. As at IG XII.5 739.7 (first cent.; cf. SH 977) or frequently in Archaic and Classical poetry (cf. LSJ, ἠπύω, 1–4).return to text

      45. Franco 2003, 161.return to text

      46. Franco 2003, 141; for other ways the Indian type of hound specifically could be a status symbol, see above, footnote 31.return to text

      47. See Bernard 2004, 231 for an inscription that more blatantly honors the dedicator in terms of his relation to his household, all within a similar structure of "episodes."return to text

      48. See LSJ, s.v. along with Sch. Il. (Sch. Vet.) X.167 and XVI.29 Erbse and Suda alpha 1588: [...] Ἀμήχανος, πρὸς ἣν οὐκ ἔστι μηχανήσασθαι (=Sch. Il. (Sch. Vet.) XV.14a Erbse). οὐκ αὐτὸς μὴ δυνάμενος μηχανήσασθαι. [...] ("ἀμήχανος: [meaning] she against whom it is not possible to contrive, not he who is not able to make a contrivance himself").return to text

      49. Note that the dog is cremated, like a member of Zenon's family would be. As far as we know, Zenon had no wife or natural-born children, a state of affairs that may have been a precondition of his service to Apollonios the dioikētēs. Contra Pomeroy 1997, 218, Clarysse and Vandorpe 1995, 61–62 point out that a Cleon who addresses Zenon as "father" (P.Cair.Zen. III 59457, IV 59580; PSI V 528) is almost certainly a war orphan taken under his care. Plut. Demetr. 25.8 suggests that, for this period at least, some financial/managerial positions were thought to be suitable for those unable or unwilling to produce children, presumably to minimize the principal/agent problem. An agent with direct descendants would have a greater incentive not to act in the interest of the principal, particularly since the agent has greater information (an "information asymmetry") about decisions made on the principal’s behalf. Franko 1988, 71–73 in fact has found instances in which Zenon manipulates his position as an agent to the profit of his brothers at Apollonios' expense.return to text

      50. Neer 2002, 87–134.return to text

      51. 21.9–10: καὶ κυναρίου δὲ Μελιταίου τελευτήσαντος αὐτῷ μνῆμα ποιῆσαι καὶ στηλίδιον ποιήσας ἐπιγράψαι 'κλάδος Μελιταῖος' ("[The man with petty ambition] makes a monument to his pet Melitian dog when it dies and when he has a little stele made, inscribes on it 'Melitian Klados' ").return to text

      52. As astutely observed by Hicks 1882, 31. Unfortunately, he follows this observation with overzealous attempts to emend the text. Though Mentz 1933, 128 notes the similarity with the name Θάλλων, which is attested for dogs, we have one attestation of this name for a person in BGU X 1943.14 (215/214; Pros.Ptol. 9802a).return to text

      Works Cited

      Bernard 2004    P. Bernard et al. "Nouvelles inscriptions grecques d'Asie centrale" Journal des Savants (Jul.-Dec. 2004) 227–356.
      Cazzaniga 1973    I. Cazzaniga, "Pap. Zenon 59532; epigramma in distici per la morte del cane Tauron," Eirene 11 (1973) 71–89.
      Clarysse 1983    W. Clarysse, "Literary Papyri in Documentary 'Archives'," Egypt and the Hellenistic World. Proceedings of the International Colloquium Leuven 24–26 May 1982, Stud.Hell. 27 (1983) 43–61.
      Clarysse and Vandorpe 1995    W. Clarysse and K. Vandorpe, Zenon, un homme d'affaires grec a l'ombre des pyramides (Leuven 1995).
      Cribiore 2001    R. Cribiore, Gymnastics of the Mind: Greek Education in Hellenistic and Roman Egypt (Princeton 2001).
      Day 1984    L.P. Day, "Dog Burials in the Greek World," AJA 88 (1984) 21–32.
      De Luca 2002    C. De Luca, "Literary and Subliterary Papyri from Philadelphia in Cairo Museum" in M. Eldamaty and M. Trad (eds.), Egyptian Museum Collections around the World. Studies for the Centennial of the Egyptian Museum (Cairo 2002) I 259–267.
      Devine and Stephens 1994    A. Devine, and L. Stephens, The Prosody of Greek Speech (New York 1994).
      Edgar 1920    C. Edgar, "Tomb-Stones from Tell el Yahoudieh," ASAE 19 (1920) 101–104.
      Edgar 1931a    C. Edgar, Zenon Papyri. Catalogue général des antiquités égyptiennes du Musée du Caire (Cairo 1931).
      Edgar 1931b    C. Edgar, Zenon Papyri in the University of Michigan Collection (Ann Arbor 1931).
      Evans 2004    T. Evans, "Orality, Greek Literacy, and Early Ptolemaic Papyri," in C.J. Mackie (ed.), Oral Performance and its Context. Mnemosyne Suppl., 248. (Leiden-Boston 2004) 195–208.
      Evans 2005    T. Evans, "Valedictory ΕΡΡΩΣΟ in the Zenon Archive Letters of Hierokles" ZPE 153 (2005) 155–158.
      Fantuzzi and Hunter 2004    M. Fantuzzi and R. Hunter, Tradition and Innovation in Hellenistic Poetry (Cambridge 2004).
      Franco 2003    C. Franco, Senza ritegno: il cane e la donna nell'immaginario della Grecia antica (Bologna 2003).
      Franko 1994    G.F. Franko, "Sitometria in the Zenon Archive," BASP 25 (1994) 13–98.
      Gortemann 1957    C. Gortemann, "Sollicitude et amour pour les animaux" CdÉ 32 (1957) 116–118.
      Herrlinger 1929    G. Herrlinger, Antike Tier-Epikedien (Tübingen 1929).
      Hicks 1882    E. Hicks, "On the Characters of Theophrastus," JHS 3 (1882) 128–143.
      Hengstl 1978    J. Hengstl, Griechische Papyri aus Ägypten: als Zeugnisse des öffentlichen und privaten Lebens (Munich 1978) 240–242.
      Hickey 2009    T.M. Hickey, "Tebtunis on the Arno and Beyond: Two 'Archives'" in G. Bastianini and A. Casanova. (eds.), 100 anni di istituzioni fiorentine per la papirologia: Atti del Convegno internazionale di studi, Firenze, 12–13 giugno 2008 (Florence 2009) 67–81.
      Johnson 1993    W. Johnson, "Pliny the Elder and Standardized Roll Heights in the Manufacture of Papyrus," CPh 88 (1993) 46–50.
      Johnson 2004    W. Johnson, Bookrolls and Scribes in Oxyrhynchus (Toronto 2004).
      Kidd 1997    D. Kidd, Aratus Phaenomena (Cambridge 1997).
      Lloyd-Jones and Parsons    H. Lloyd-Jones and P. Parsons, Supplementum Hellenisticum (Berlin 1983) 977.
      Loraux 1995    N. Loraux, The Experiences of Tiresias, trans. P. Wissing (Princeton 1995).
      Merkelbach and Stauber 1998    R. Merkelbach and J. Stauber, Steinepigramme aus dem griechischen Osten. 5 vols. (Stuttgart 1998–2004).
      Mentz 1933    F. Mentz, "Die klassischen Hundenamen" Philologus 88 (1933) 104–129, 181–202, 415–442.
      Neer 2002    R. Neer, Style and Politics in Athenian Vase-Painting: The Craft of Democracy, ca. 530–460 B.C.E. (Cambridge 2002).
      Orrieux 1983    C. Orrieux, Les papyrus de Zénon (Paris 1983).
      Orrieux 1985    C. Orrieux, Zénon de Caunos (Paris 1985).
      Page 1981    D. Page, Further Greek Epigrams (Cambridge 1981).
      Page 1942    D. Page, Greek Literary Papyri (Harvard 1942).
      Pomeroy 1997    S. Pomeroy, Families in Classical and Hellenistic Greece: Repre–sentations and Realities (Oxford 1997).
      Purola 1994    T. Purola, "P.Cairo Zenon 4.59532 – Two Epitaphs for a Hunting Dog Called Tauron," Arctos 28 (1994) 55–62.
      Radermacher 1951    L. Radermacher, Artium scriptores (Reste der voraristotelischen Rhetorik. SAWW 227.3 (Vienna 1951).
      Reichel and Wilhelm 1901    W. Reichel and A. Wilhelm, "Das Heiligtum der Artemis zu Lusoi," JÖAI 4 (1901) 1–89.
      Robert 1960    L. Robert, Hellenica XI–XII (Paris 1960).
      Saija 1990    A. Saija, "Considerazioni metriche su P. Cair. Zenon 59532: epigrammi per il cane di Zenone," Eirene 27 (1990) 29–34.
      Sundqvist 2006    A.–K. Sundqvist et al. "Unequal Contribution of Sexes in the Origin of Dog Breeds," Genetics 172 (2006) 1121–1128.
      Viereck 1928    P. Viereck, Philadelphia: die Gründung einer hellenistischen Militärkolonie in Ägypten. Morgenland 16 (Berlin 1928).
      Wilhelm 1900    A. Wilhelm, "Der Dichter Antiphon aus Athen," JÖAI 3 (1900) 93–98.