Scroll for More: Papyrus Commentaries from Roman Oxyrhynchus and Context Clues

1 Reuse

A first possible clue about the circumstances of a copy’s use is whether it was made on good quality papyrus or on reused material.^[7] Turner 1956 lists reuse as a quite typical feature of scholars’ texts.

We have examples of commentaries on the back of all kinds of texts. Sometimes they are paraliterary:

• P.Oxy. II 221 comm. on Il. 21 (2nd c. AD) on the back of P.Oxy. II 220, a treatise on metrics (2nd c. AD)

• P.Oxy. LXXX 5231 comm. on Hp. Epid. 1 (1st/2nd c. AD) on the back of a bilingual glossary P.Oxy. LXXVIII 5162 (1st/2nd c. AD)

Sometimes the commentary is copied on the back of documents:

• P.Strasb. P. gr. inv. 84^[8] comm. on Demosth. Contra Androtionem (1st/2nd c. AD) on the back of an account (1st c. AD)

• P.Oxy. XXVI 2451v comm. on Pind. Isth. (1st/2nd c. AD) on the back of an official account book (mid. 1st c. AD)

• P.Oxy. LIV 3722 comm. on Anacreon (2nd c. AD) on the back of a month-by-month account (1st/2nd c. AD)

• P.Oxy. XXVII 2463 (?) fragment of a learned discussion on the myth of Poemander possibly from a comm. on Lyc. Alexandra or Call. Aitia (2nd/3rd c. AD) on the back of a tax register

Sometimes they are even written on the backs of a pasted roll, that is to say a composite set of documents specially pasted together to provide a roll of sufficient length:^[9]

• P.Oxy. LXV 4452^[10] comm. on Il. 19 (2nd c. AD)

• P.Oxy. VI 853 comm. on Thuc. 2 (2nd c. AD).

Commentaries on reused material are often informal but professional copies, perhaps custom-made products or private copies by a person with a good scribal training.

Of course, tracing a link between the text on the recto and that on the verso is difficult, and speculation regarding it may be pointless. These rolls may have been written elsewhere and then taken home by officials returning from a period of service, may have been stored somewhere and then reused after some time by unrelated persons, or may even have been bought second-hand as paper for recycling.^[11] Yet, with due caution, reuse is undoubtedly worth considering, even if a full comprehension is often hampered by the fact that some of the texts on the other side have merely been described and are still unedited.

The same considerations hold true, conversely, for the cases where the commentary bookroll was reused on its reverse^[12] to receive a documentary text:

• P.Ryl. I 24r comm. on Il. 4 (1st c. AD) reused for a Demotic account (1st c. AD)

• P.Oxy. XXX 2528r comm. on a poem (Euphorion?) (in. 2nd c. AD) reused for a land register (2nd/3rd c. AD)

• PSI inv. 90 comm. on poetry (Od. 23?) (2nd/3rd c. AD)^[13] with remains of cursive script on the verso (2nd c. AD)

• PSI XIV 1391 comm. on Pindar? (mid. 2nd/early 3rd c. AD) with a cursive list of food items on the verso (mid. 3rd c. AD)

Sometimes a commentary roll receives a paraliterary text on its back:

• P.Oxy. XIX 2221+P.Köln Gr. V 206r comm. on Nicander^[14] (1st c. AD) reused for a medical text (1st c. AD)

• P.Oxy. XVII 2086r comm. on comedy (2nd c. AD) reused for a rhetorical treatise with quotations from Demosthenes (3rd c. AD)

Sometimes such a roll is even reused for a good quality copy of a literary text, as in the case of the learned commentary on Il. 1 (late 1st/early 2nd c. AD) from the Bodleian Library (MS Gr. class. f. 110r) latterly reused for a copy of Il. 4 in ‘severe style’ (3rd c. AD).^[15]

These examples may provide some evidence on how long such products were valued (how much time passed before the primary text was repurposed?), yet in some instances the two texts seem to be contemporary or nearly contemporary, so the presumed reuse could actually be a concurrent use of both sides. Even when a long time passed between the two texts, one may suspect that the commentary may still have been of interest to someone who copied a literary or paraliterary text on the back of it.^[16]

2 Repair Patch

A more reliable indication of a long-lasting interest in a commentary roll may be the presence of repair patches,^[17] so long as we assume that the repairs were made after the text had been copied and used for some time, rather than beforehand to reinforce a still blank roll. An example is P.Berol. inv. 13413, a copy of a commentary on the second book of the Argonautica in a well executed bookhand datable to the first or second century AD (CLGP Apoll. Rh. 6), though its provenance from Oxyrhynchus is uncertain.^[18] Haslam noted the presence of a couple of scraps of documents on the back of it, attached to reinforce the roll (“This may be evidence that the commentary was sufficiently valued as to be patched up when it began to disintegrate” Haslam 2011, 30). In one of these patches the addressee is Lucretius Nilus, who was strategos of the Oxyrhynchite nome in AD 197–198 and who had previously held offices in the Arsinoite.

3 Identified Scribe

We have some ten copies of commentaries by the hand of identified Oxyrhynchus scribes who produced other bookrolls.^[19] The identification of a roll as the product of a given scribe is occasionally disputed, but the specific graphic style may still point to a common origin and may provide insight into the preferences and interests of the clientele for such products. It seems a sound assumption that scholars – the most likely owners of commentaries – would have trusted and frequented particular copyists as loyal customers.^[20]

Known Oxyrhynchus scribes held to be responsible for copies of commentaries are:

The hand of scribe A5^[25] (late 2nd/early 3rd c. AD according to Johnson 2004, first half of 3rd c. according to Funghi/Messeri 1992a, 76) has been identified in three fragments of commentaries on poetic texts, two of which were later reused on the back for documents (see supra).^[26] The same scribe also wrote bookrolls of Herodotus 2 with textual notes (P.Oxy. VIII 1092) from the second large find by Grenfell and Hunt (see infra), Plato’s Phaedo (P.Oxy. LII 3676), Alcaeus (P.Oxy. XXI 2297, with peculiarly wide intercolumns of 5 cm), Alcman (P.Oxy. XLV 3213), the Hellenistic poetry of Euphorion (PSI XIV 1390), and a historical work (P.Mich. inv. 4913). All products were executed on very good quality papyri.

Lobel argued that a single scribe produced more than one commentary also in the cases of scribe A18, identified in the three hypomnemata on Alcaeus, Bacchylides and a comic adespoton,^[27] and scribe A19, who, besides the commentaries on Alcman and on Iliad 17, also produced elegant copies of poetic works,^[28] namely Simonides’ Elegies (P.Oxy. XXII 2430 and P.Oxy. XXII 2327, both with marginal notes), iambic trimeters of unknown authorship with a few traces of interlinear notes (P.Oxy. XXII 2318 = CLGP Archilochus 6 [?]) and perhaps also a copy of Apollonius Rhodius (P.Oxy. XXXIV 2694),^[29] i. e. books for a very learned clientele. The poetic rolls have large intercolumns and bear annotations possibly by the same hand. The same is perhaps true of the few notes in the Alcman commentary P.Oxy. XXIV 2389, too, but the remains of the notes are too fragmentary to be certain.^[30]

4 Secondary Interventions

Another contextual clue is the possible presence of interventions by a second hand.

Many copies of commentaries were checked by a diorthotes. But as well as simple corrections and sigla that marked points to be checked, there are also cases of marginalia containing additional materials^[31] and attesting to an ongoing scholarship and sometimes perhaps indicating a working copy. A possible example of the latter is P.Oxy. LIV 3722 (2nd c. AD),^[32] a learned commentary on Anacreon penned on reused material and presenting various additions, corrections and improvements by the first hand.

In the case of P.Oxy. XIIX 2176 (1st/2nd c. AD), a fine copy of a good level commentary on Hipponax (with mention of grammarians) in an elegant informal bookhand,^[33] the text underwent an attentive revision and the corrector also added cursive notes, both interlinear and marginal, with additamenta of various kinds: prosopographical information, metrical, grammatical and lexicographical notes, and paraphrase. The hand seems contemporary to that of the main text.^[34]

Three interlinear and marginal notes by a second hand, two of them citing Didymus, are preserved in P.Flor. II 112 (2nd c. AD), a commentary on an unknown play by Aristophanes, which according to Lobel was copied by the same scribe (A15) as a copy of Alcaeus with marginal notes (P.Oxy. XXI 2301).^[35]

Additions are also present in P.Oxy. XXXI 2536, a commentary on Pindar’s Pythians by Theon, to which I will return later.

Furthermore, meagre remains of marginalia or simply sigla are preserved in:

• P.Oxy. XXIV 2389 comm. on Alcman (1st c. AD) (scribe A19)

• P.Oxy. XXIII 2368^[36] comm. on Bacchylides (2nd c. AD) (scribe A18)

• P.Oxy. XXXII 2637^[37] comm. on lyric poetry (Ibycus?) (2nd c. AD)

• P.Oxy. LXV 4452^[38] comm. on Il. 19 (2nd c. AD) (on reused material)

• P.Oxy. XXVI 2449^[39] comm. on Pindar? (2nd/3rd c. AD)

• P.Oxy. XXXV 2741^[40] comm. on Eupolis’ Marikas (2nd/3rd c. AD)

• P.Oxy. LIII 3710^[41] comm. on Od. 20 (2nd c. AD) (scribe A5)

Sometimes it is unclear whether an annotation is related to the text of the commentary or not. The case of the intercolumnial annotation in P.Oxy. II 221, an extremely learned commentary on Il. 21 of the second century AD, is well known and much debated. It reads “Ammonios son of Ammonios grammatikos ἐϲημειωϲάμην”.^[42] What does ἐϲημειωϲάμην mean here? And who is this Ammonios?

He made a ϲημεῖον (‘sign’): Does this mean critical signs? Or a signature? In documents ϲεϲημείωμαι means ‘I have put my seal’ i. e. ‘I have certified/approved’ a given document (cf. Preisigke). The most famous Ammonios who engaged in Homeric exegesis is the well-known pupil and successor of Aristarchus, and it would be no surprise to find a commentary under his authority in Roman-period Oxyrhynchus, even though the text cites later Augustan- and Tiberian-period grammarians such as Didymus, Aristonicus, and Seleucus (but not Herodianus). It is not strange at all that, starting from an authoritative foundation, a commentary may develop in a quite open or fluid tradition, through additions, epitomizations and alterations.^[43] Undoubtedly, a famous name (properly claimed or not) could lend the work more authority.^[44] But if the Ammonios at issue is actually the second-century BC Alexandrian grammarian, why the first-person ἐϲημειωϲάμην – was it some kind of emulative mimicry?

5 Colophon or Other Indication of Authoriality

If we set aside the peculiar case of P.Oxy. II 221, there are some rolls that preserves the colophon with indication of the author of the hypomnema.^[45] Besides the famous cases from Hermopolis – Aristarchus on Herodotus (P.Amh. II 12) and Didymus on Demosthenes (P.Berol.)^[46] – from Oxyrhynchus we have two:

• P.Oxy. XXXI 2536 comm. on Pindar by Theon;

• P.Oxy. XXIV 2392 comm. on Alcman’s carmina book 4 by Dionysius (2nd c. AD).^[47]

The identity of Dionysius is very controversial, but Theon is quite well known.^[48] He was the son of Artemidorus, the influential grammarian known for his exegetical and lexicographical activities on various poets and thought to have had specific connections with Oxyrhynchus.^[49]

The surviving part of the commentary, on Pythian 12, provides mainly mythographical information (on the figure of Perseus) and linguistic notes, including a quotation from Euripides’ Orestes. It is likely an excerpted version of Theon’s original commentary, though it is more extensive than the information under the same lemmata in the scholia. Worthy of note is that the compilation was not written by a single scribe. There are at least two different hands and the copy has marginal annotations. The first editor, Turner, distinguished three hands (hand 1: col. I 1–26; hand 2: marginalia; hand 3: col. I 27–30 and col. II), and McNamee also argued for additions by two later users. More recently, Ucciardello has argued that the second and third hands may well be the same, that this one hand may also be responsible for the end-title, and that both scribes worked together at the same time (second half of 2nd c. AD). Apparently, they had different interests and their work was aimed at different levels in a sort of “collective transcription of Pindar exegesis”: the first copyist-reader has a mainly mythographic focus, while the second one is more sophisticated.^[50]

Additionally, although only the genitive ending -ου is legible, the putative author of a hypomnema on Aristophanes is likely to have been present also in one of the two title tags for commentary rolls.^[51]

6 Found Together

All these authoritative reference-works eventually ended up in the rubbish.

Regrettably, the archaeological information about the circumstances of discovery of papyri is often completely lost. As is well known, for Oxyrhynchus we only rarely have precise information on findspots or stratigraphic levels, and the archeological reports are poor and frustratingly hazy. Yet, when we do have some information, proximity to other documents from the same findspot may provide useful clues. Admittedly, the proximity of two items in the disposal site is not necessarily significant, since it could be due to sheer chance and, in the archaeological context of a dump, contaminations are a distinct possibility. Nonetheless, the discovery of concentrations of literary papyri in a specific site at the rubbish dumps of Oxyrhynchus has, ever since Grenfell and Hunt, led to the assumption that these papyri could be part of a single ancient library that was thrown away.^[52] And indeed this assumption seems reasonable, though impossible to establish with certainty. During their fifth season of excavation, Grenfell and Hunt were able to “realize the long-cherished ambition” to discover “classical works from a scholar’s library”,^[53] as they announced, making three great literary finds in the winter of 1906. Some years later, in spring 1932, another very large concentration of literary papyri at Oxyrhynchus was discovered by the Italian mission headed by Evaristo Breccia.

The first group is the so-called Hypsipyle library: “a basketful of broken literary papyrus rolls thrown away in the III AD”, found on 13–14 January 1906. The basket^[54] seems to have included 16 books, all datable to the second or third century AD, of various genres: history, oratory, philosophy (Plato) and poetry (Euripides, Pindar).^[55] Almost half of these books are written on reused material, a striking proportion.^[56] Despite being on reused material, the books appear to be professional products.

From this collection we have one of the instances of a hypomnema written on the verso of a pasted roll: P.Oxy. VI 853, a substantial commentary on Book 2 of Thucydides’ history (2nd c. AD).^[57]

The three documents glued together to form the roll (P.Oxy. VI 986) are:

1. an administrative report by the komogrammateus dated to the 16th year of Hadrian (131–132) concerning properties and land confiscated in Oxyrhyncha (Arsinoites);

2. a return by sitologoi to an officer of the Polemonos meris;

3. a seed account.

These documents originate in the Arsinoite nome but ended up in Oxyrhynchus – whether before or after they were recycled cannot be determined. A parallel example is the long Land-Survey Register P.Oxy. VI 918 (listing land plots in the meris of Polemon), which was reused for P.Oxy. V 842, the famous Hellenica Oxyrhynchia. Probably these official papers were taken home to Oxyrhynchus by someone who had held an administrative position in the Arsinoites.

The informal script of the commentary is likened by the editors to that of the so-called Commentary of Ammonios on the Iliad, P.Oxy. II 221.

The same library presumably included copies of the author commented upon. In the basket there was also a roll of Book 7 of Thucydides (P.Oxy. XI 1376, 2nd–3rd c. AD). McNamee proposed a possible attribution to this group of some other books, on the grounds that they were written by the same scribes: these include the fragments of the first five books of Thucydides in a codex in ‘severe style’ by scribe A35 (the same responsible for P.Oxy. VII 1016, Plat. Phaedr., both mid-3rd c. AD).

Some of the books in this group have marginal notes: a copy of Plato’s Symposium (P.Oxy. V 843), the roll of Euripides’ Hypsipyle (P.Oxy. VI 852) that gives the archive its name, and a copy of Pindar’s Paeans (P.Oxy. V 841). The latter two rolls were professionally executed on reused material, as suggested by script, layout and the presence of stichometric counts.^[58] The Pindaric roll has extensive annotations by at least three hands, one of which is probably the scribe of the main text. These added both glosses and explanatory notes, recorded variae lectiones, sometimes attributing them to specific persons,^[59] and marked some poetic passages with chi, diple and ζή(τει). The names cited in the notes possibly include Theon, who is cited also in the margins of other bookrolls of poetry (Epicharmus, Sophocles’ Ichneutai, an earlier copy of Stesichorus) that likewise have multiple annotators who attach a name to variant readings.^[60] These annotated books attracted the attention of Johnson 2010, who described them as vivid evidence of “a scholarly behaviour in which the reading and study and analysis of difficult texts is constructed as a collective endeavor”, a group-work by the social elite “play(ing) to be intellectuals” (p. 192). Interestingly enough, the work on the aforementioned copy of the Pindaric hypomnema by Theon is also “collective”, providing additional support for Johnson’s thesis regarding a collective endeavour via a second-century community of readers engaged in the study of Pindar’s poetry at Oxyrhynchus. We cannot know, of course, whether this is the same group that was perusing the authoritative commentary of Theon for the analysis of Pindar’s Paeans. It seems not impossible, after all, that the same people could be interested both in copying extracts from the hypomnemata by Theon (P.Oxy. XXXI 2536) and in taking notes of the results of this kind of commentaries and other scholarly products on the bookrolls of Pindar’s poetry they were reading (P.Oxy. V 841).

There is no firm evidence for the presence of commentaries in the second large literary find Grenfell and Hunt made – the so-called Ichneutai Library discovered in January–March 1906 in the mound known as Kom Ali el-Gamman. Among other poetic rolls it yielded the eponymous roll of Sophocles, which includes multiple annotators citing Theon, among other names (P.Oxy. IX 1174 + XVII 2081(a)). Remarkably, almost half the books in this find have annotations (45 %, against a general estimate of 5 % by McNamee 2007a) and there seems to be a strong interest in poetry, especially lyric, in what appears to be a collection of scholars’ books.^[61] The books range considerably in date (mainly 1st to 3rd c. AD)^[62] but were discarded far later (about AD 400). McNamee 2007b suggested that some commentaries should also be included in this group, on the grounds that they were written by the same scribes as rolls certainly belonging to it:

• P.Oxy. LIII 3710, comm. on Od. 20 (mid 2nd/early 3rd c. AD) by the same scribe A5 as a Herodotus copy in this group (P.Oxy. VIII 1092)^[63]

• P.Oxy. XXI 2307, comm. on Alcaeus (CLGP 12) (2nd c. AD) by scribe A32.

The Alcaeus case is especially intriguing. According to Haslam, the commentary is by the same scribe (Oxy. scribe A32)^[64] as the copy of Alcaeus from the second large find by Grenfell and Hunt (P.Oxy. X 1233).^[65] We could here have the pair: the roll of the work commented upon and the roll of the commentary.^[66] In both rolls a second, very similar hand added some marginal notes.^[67]

Also uncertain is the possible inclusion of two commentaries in the third group of literary papyri found by Grenfell and Hunt in 1906, which was from “another part of the same mound” Kom Ali el-Gamman:^[68]

• P.Oxy. XVII 2085, comm. on Euphorion (1st/2nd c. AD)^[69]

• P.Oxy. XVII 2086r, comm. on a comedy of unknown authorship (2nd c. AD) (CLGP Adesp.Com. 8)

Attribution to this group depends solely upon the prefatory remarks in the volume in which they were published. In the preface to P.Oxy. XVII, p. v, Hunt says that “many” of the literary texts in the volume come from the third large group found in January 1906, a perfect illustration of the discouraging haziness of the information available.

Besides the three great finds by Grenfell and Hunt,^[70] two commentaries survive from Breccia’s find of spring 1932 in Kom Ali el-Gamman (published in PSI XI and XII), which was named for the medieval sheikh whose tomb lay atop it. The tomb’s presence prevented Grenfell and Hunt from investigating the mound thoroughly, but Breccia resumed work after the Italians relocated the tomb. There, they retrieved many literary fragments together with documents which included part of the papers of the strategos of the Arsinoite and Hermopolite nomes, Sarapion alias Apollonianos, and his family.^[71] The fascinating hypothesis that the literary rolls belong to his library remains speculative (Funghi/Messeri 1992, 77). Funghi and Messeri argued that this find of 1932 may be from the same library as Grenfell and Hunt’s second large find, because it came from the same mound and the digging was at the same depth, and because the hands at work in some documents from the two concentrations are the same (scribe A5, among others). To the contrary, Houston 2007, adopting the same methodology but reaching different conclusions, believes it is the same concentration as Grenfell and Hunt’s third group, not the second one (cf. also Houston 2014, 143–158), and the issue is still debated.^[72]

The literary rolls from Breccia’s 1932 find include fifth-century dramatic works (e. g. Eupolis PSI XI 1213), Hellenistic poetry (Callimachus, Sophron, Euphorion), oratory, history and the novel (Ninus romance).

The find also yielded a pair of commentaries:^[73]

• PSI XI 1219, comm. on Callimachus’ Aitia 1 (2nd c. AD), perhaps a fragment from the beginning of the roll.^[74] There are three rolls of Callimachus’ Aitia among the papyri from Breccia 1932 (Houston nos. 8, 9, 10), so much so that this could be another potential case where we have the literary text and the associated hypomnema.

• PSI XIV 1391,^[75] comm. on Pindar (= Pind. fr. 346 Maehler) by the scribe A5 (mid-2nd/early 3rd c. AD) reused on the back for what seems to be a list of food items. This papyrus thus has three of our “context clues” (identified scribe, reuse, findspot in a concentration).

Conclusions

This overview pursuing intertwined clues leads to unsurprising conclusions.

The discovery of commentaries among concentrations of literary papyri, and the examples by identified scribes, tell us that commentary rolls were not alone on the shelf: they were produced and used together with books of literature and other commentaries.^[76] They were typically part of book collections, often – so it seems from the informal copies on reused materials – the private collections of serious readers. These serious readers were cultivated men with an above-average interest in classical literature and traditional philology, not necessarily professional scholars,^[77] perhaps teachers, but they were certainly playing an active part in the bustling intellectual life of Oxyrhynchus.^[78]

The presence of interventions in some of these papyri by more than one hand argues for the circulation and production of these copies in reading circles, which shared a strong drive to understand and get to grips with the difficult texts of the Greek literary heritage, using and perhaps continuing the authoritative philological tradition of Alexandria.