3.1 The Typology of the Documentary Texts on the Recto

The first aspect to consider is the typology of the documents on the recto. Lama proposes a distinction between official documents that were preserved in public archives and private documents.^[20]

A preliminary problem is that all the documents on the front of the volumina bearing Menander’s comedies on the verso are currently unpublished. In the absence of a direct examination, we have only the scant information provided in the edd. pr. of the literary fragments.^[21]

On the basis of the available data, the following papyri are likely to preserve official documents on the recto: P.Oxy. XLI 2943 [8], which bears a fragment (possibly from a register) of official correspondence;^[22] P.Oxy. XLVIII 3368 [12], containing a register recording payments of a local tax for the month of Thoth (year unknown) at a place in the Oxyrhynchite nome; and P.Oxy. LX 4021 [14], which bears mutilated remains of three documents from a tomos synkollesimos.

In the other twelve cases we are unable to determine the content with certainty. Of the documents on the recto of P.Oxy. LX 4024 [1], P.Oxy. LXXIII 4936 [6], P.Oxy. LXXIX 5198 [9] and P.Oxy. LX 4019 [13], we know only that remains of cursive survive, which prevents an assessment of the contents. The same is true of P.Oxy. L 3533 [4], which, according to Turner, bears a Latin register, but only a few letters are legible.^[23] P.Oxy. XLVIII 3370 [3], P.Oxy. LIX 3967 [5] and P.Ross.Georg. I 10 [7], currently lost, preserve on their rectos an unspecified type of account. Similarly, there is nothing to indicate either an official or a private purpose for the registers on the front of P.IFAO inv. 89v + P.Köln VII 282 [10] or P.Oxy. LXIV 4407 [15]. At the moment, nothing can be said about the document on the recto of the Michigan roll, bearing on the back a copy of the Epitrepontes [2], that was discovered at Karanis in 1926 and probably belonged to Socrates, son of Sarapion, πράκτωρ ἀργυρικῶν in the village in the first half of the 2nd c. AD.^[24] Finally, we do not know if the contract that is preserved on the recto of P.Oxy. XXXIII 2657 [11] is a chirograph, an agoranomic contract intended for registration, or even the private copy of such a document.

3.2. The Chronological Relation between the Use of Recto and Verso

In all the cases considered, the literary text on the verso was likely written when the documentary text on the recto had already fallen into disuse. It is thus of undoubted interest to establish, where possible, how much time elapsed between the writing of recto and verso.

One obstacle is the almost total lack of chronological indication on the papyri under examination. Not only is this true for the literary texts on the versos, which can be dated only on palaeographical grounds (as would be expected), but also for the documents on the recto, whose dates do not survive or have yet to be read.

Moreover, as it is well known, the circumstances of preservation and reuse of documents is a controversial topic. For private documents these aspects are largely imponderable, as they may depend on personal factors such as the interests, wishes and organisational habits of the individual owner, as well as on the typology and practical usefulness of a particular document.^[25] On the other hand, very different proposals have been advanced for the storage of official documents, without reaching any definitive conclusions.^[26]

Among those considered here, only three documents offer chronological insights.

The first is preserved on the recto of the Michigan roll bearing the Epitrepontes on the back [2]. It can be dated to the reign of Domitian on the basis of the first line of P.Mich. inv. 4733, fr. 2 recto.^[27] Michael Gronewald has convincingly assigned the literary copy on the verso to the beginning of the 2nd c. AD, in line with the fragment’s possible origin as part of the archive of Socrates.^[28]

P.Oxy. LX 4021 [14] bears on the front scanty remains of three documents from a tomos synkollesimos. Clearly legible on l. 15 of fr. 1 is a portion of a dating-clause (]Αὐρηλίου Ϲεο[υήρου]) referring to the sole reign of Caracalla (AD 211–217) or the reign of Alexander Severus (AD 222–235). The literary text on the back is written in a slovenly script with cursive tendencies that is datable to the 3rd c. AD (perhaps the first half).^[29]

The reuse of these two rolls seems to have occurred in close proximity to the copying of the recto texts – at most, within a few decades. This is also clearly the case with P.Oxy. LXIV 4407 [15], which preserves on the recto “a species of register” from AD 241/242.^[30] The literary text on the back is written in a professional, but not elegant, ‘severe style’, rightly assigned to the end of the third or beginning of the 4th c. AD by Eric W. Handley.^[31]

The documents on the recto of P.Oxy. LXXIII 4936 [6], P.Ross.Georg. I 10 [7], P.Oxy. XLI 2943 [8] and P.Oxy. XLVIII 3368 [12] are dated on palaeographical grounds too generically to draw specific conclusions.^[32]

3.3 Features of the Literary Copies

Let us now turn to the features of the literary texts written on the document versos. In this respect, palaeographical characteristics are undoubtedly the first element to consider. Lama distinguished between papyri in calligraphic script, papyri in documentary script and scholastic papyri.^[33]

Leaving aside the controversial category of scholastic papyri, I would like to briefly consider the quality of copies of Menander made on the verso of documentary papyri, starting with the features of the handwriting. I will employ the categories delineated by Guglielmo Cavallo in a 2013 paper in the Proceedings of the 17^th International Congress of Papyrology, which correspond better to the variety of graphic manifestations that has emerged from the survey conducted here. Cavallo assigned the scripts to four categories distinguished by the form of the letters and by the ductus:

1. ‘normative or formal scripts’, characterised by careful and sometimes calligraphic traits, a slow and clearly legible ductus, most often identified by precise nomenclature in modern palaeography;

2. ‘semi-formal scripts’, written with less controlled ductus that simplifies aspects of the letters, which nevertheless remain clear, and with a certain inelegance;

3. ‘informal semi-cursive or cursive scripts’, rapidly written, in which modification of the strokes within the letters and deforming ligatures result in a reduced level of legibility;

4. ‘slow informal scripts’, written with a slower ductus compared to that of the third category, but in which the forms of the letters are nonetheless modified in a semi-cursive or cursive form.^[34]

According to Cavallo, ‘normative and formal scripts’ were usually employed in professional literary copies, ‘informal semi-cursive or cursive scripts’ in private ones, while the ‘semi-formal’ as well as the ‘slow informal’ scripts could be used both professionally and privately, in the first case for copies of a secondary quality.^[35]

The palaeographical features of the Menander’s copies written on the verso of documentary texts vary: they include exemplars of very modest quality, which lack any attempt at elegance or even of order and are written instead in overtly informal or even semi-cursive or cursive script. These items are to be assigned to Cavallo’s third and fourth groups. But most Menander papyri are in semi-formal hands, belonging to Cavallo’s second group. Among these, as we shall see, there are numerous copies written in ‘severe style’, of varying palaeographical and bibliological levels, though never particularly calligraphic. Papyri in ‘normative scripts’ are completely absent (see Table 2).

The writing of P.Oxy. XXXIII 2657 [11] (Figure 1) is a good example of Cavallo’s third category, namely ‘semi-cursive or cursive scripts’. P.Oxy. XXXIII 2657 [11] preserves ll. 401–492 and fr. 3 of the Misoumenos. It is written in a rapidly executed cursive removed from any desire for elegance or even for clarity. The line spacing is irregular and the intercolumnar spacing extremely small; the copy would have had columns of ca. 50 lines and a height not less than 33 cm. The fibres have probably been already damaged at the time of reuse, as is clearly shown by the ca. 1 cm of vacant space avoided by the scribe in the body of the text in the first column. The scribe copied the text without particular care, but at times corrected it himself (see e. g. col. i 23). There are paragraphoi, dicola (in mid-line only) and occasional diastolai.

Figure 1:

P.Oxy. XXXIII 2657. Menander, Misoumenos. Courtesy of The Egypt Exploration Society and the Faculty of Classics, University of Oxford.

The papyri I assign to Cavallo’s fourth category, i. e. ‘slow informal scripts’, have features similar to those of the third group. A good example is P.Oxy. LX 4024 [1], the oldest papyrus in which Menander’s text figures on the verso. It is datable to the 1st c. AD and preserves ca. 10 lines from a Menandrian comedy, probably the Leukadia, written in “a graceless informal hand”.^[37] Changes of speaker are marked by paragraphoi and dicola, there is an interlinear correction (l. 3) and possibly a variant (l. 4), the former apparently written by a different hand in a darker ink.

Due to their overall characteristics, it seems reasonable to conclude that these fragments are from private copies, probably written in the owners’ hands by persons interested in the texts copied. Despite their very poor quality, both of the papyri briefly described here are attentive to the author’s text, as proved by the presence of stage indications, corrections, and the aforementioned possible variant in l. 4 of P.Oxy. LX 4024 [1].

The copies in ‘semi-formal scripts’ (Cavallo’s second group) present rather varied features.

P.Oxy. LXXII 4936 [6] (Figure 2), which preserves Epit. 41–109 in two consecutive columns, is of a very modest quality. The writing is small and professional-looking but appears not particularly calligraphic. The extremely small intercolumn, a certain disorder in the arrangement of the text, with the notae personarum placed close to the end of the lines of the preceding column, and the tight and irregular interlinear spacing give the impression of a crammed copy. Some phonetic spellings occur (see col. i 21 and 22) and elision is not systematically marked by apostrophe. Nevertheless, the scribe has paid attention to changes of speaker, marked by paragraphoi and marginal callouts, and in col. i 16 there may be a correction, if the scanty traces legible here are not the remains of a mid-line nota personae.

A slightly more careful specimen is the copy of the Epitrepontes found at Karanis [2] (Figure 3), repeatedly mentioned above. The informal, round script is legible but inelegant; the text is arranged in columns of ca. 38 lines, thus the copy presents a manageable format, with a low height (ca. 24 cm) compared to contemporary or slightly later rolls from Oxyhrynchus that are written on the verso.^[38]

Figure 2:

P.Oxy. LXXII 4936. Menander, Epitrepontes. Courtesy of The Egypt Exploration Society and the Faculty of Classics, University of Oxford.

Figure 3:

Fragments of the Michigan roll of Menander’s Epitrepontes. Courtesy of the Michigan University Library.

The text preserved by the papyrus overlaps with other copies of the play and attests singular readings. The scribe makes numerous errors, the notation of ‘mute ι’ is not constant, nor is the treatment of elision. The changes of speaker are indicated by dicola and notae personarum inserted by the main hand into the intercolumnar space and the body of the text.

This copy, of modest quality, is of interest in several respects. First of all, it represents one of the rare papyri bearing Menander’s comedies found outside the Oxyrhynchite nome. Moreover, if, as seems probable, it did belong to Socrates, then, in line with other papyri from sites other than Oxyrhynchos, it would document the enjoyment of this author by individuals of quite a high social and cultural level.^[39] We may also note that, in the personal archive of Socrates as reconstructed in the scholarship,^[40] Menander would have been read along with Homer and Callimachus, but also grammatical texts, the Acta Alexandrinorum, and other materials that are difficult to classify.^[41]

P.Oxy. XLVIII 3368 [12] (Figure 4) consists of numerous pieces and scraps reassembled into two large fragments (A and B) written on the verso against the fibres on a surface of poor quality. It preserves the opening lines (1–19, 33–45, 51–68, 85–100) of the Misoumenos in two consecutive columns; a large section of unwritten papyrus, probably part of the initial agraphon, precedes them. Another fragment (C) bears ll. 241–248. The roll was trimmed and probably repaired before being reused. The literary copy on the verso was arranged in columns of ca. 50 lines and would have had a considerable height (min. 33 cm). The writing, datable to the 3rd c. AD, is a medium-sized, rapidly written majuscule, frequently executed with deforming ligatures.

Figure 4:

P.Oxy. XLVIII 3368. Menander, Misoumenos. Courtesy of The Egypt Exploration Society and the Faculty of Classics, University of Oxford.

The scribe makes phonetic and metrical errors, and sometimes corrects himself (e. g. fr. C l. 8). Changes of speaker are marked by paragraphoi and occasional blank spaces (there are no dicola); there is one marginal nota personae and a possible interlinear one. The scribe rarely inserts ano stigmai; normally he uses scriptio plena (an exception is marked by apostrophe). The text of P.Oxy. XLVIII 3368 [12] overlaps with that preserved by other papyri and by fragments in indirect tradition but transmits many unique readings.^[42]

On the whole, this copy of the Misoumenos, one of the author’s most popular comedies, written on the verso of an official document, fits well with the informal character of Menander’s papyrological tradition in this period.^[43] However, the writing, which betrays a desire for clarity (despite its rapidity), the layout (with wide margins and a generous intercolumn framing fairly airy columns of text), and the presence of the initial agraphon reveal some sort of scribal plan in the copying. It is worth noting that on the recto there are also strips of papyrus pasted as a strengthening, which could be a sign of the longevity of this copy.

Finally, it is not surprising that several papyri bearing Menander’s comedies on the verso of documentary texts are written in the ‘severe style’.^[44] This is the case for P.Oxy. LXXIX 5198 [9], P.IFAO inv. 89v + P.Köln VII 282 [10], P.Oxy. LXIV 4407 [15] and, probably, P.Oxy. LX 4019 [13].

Let us discuss two cases of particular interest.

The first, P.IFAO inv. 89v + P.Köln VII 282 [10], preserves the opening lines of the Misoumenos. It is written in a quite clumsy ‘severe style’ and seems datable to the 3rd c. AD. There are numerous phonetic spellings and several corrections are made by the first hand. Elision is sometimes present, occasionally marked by apostrophe; punctuation is limited to ano stigmai. The papyrus attests some correct readings and some inferior variants that survive in the later indirect tradition.^[45] Turner posited a school use for this text, which in his opinion belonged to an anthology. Raffaella Cribiore agreed and classified P.IFAO inv. 89v + P.Köln VII 282 [10] among the school items bearing ‘long passages: copies and dictation’.^[46] According to the positions held by these two scholars, elements pointing in this direction are: the choice of the verso of a document for the copy; the graceless writing (Hand 3 in Cribiore’s categorisation);^[47] the large number of misspellings, omissions and corrections in such a short section of text (compatible with dictation?).

The theory of a school use is interesting and consistent with the popularity of the ‘inverse’ παρακλαυϲίθυρον that opens the Misoumenos.^[48] Nonetheless, given the limited evidence, the argument is inconclusive. As extensively discussed in this paper, numerous copies of Menander are written on the back of documents. The handwriting of the papyrus is certainly clumsy, but no more than those used for other copies of Menander. No other excerpts survive of the anthology hypothesised by Turner as the origin of the fragment. We would also have to assume that, in this anthology, the beginning of the excerpt from the Misoumenos coincided with the very beginning of the surviving column (which is complete at the top and preceded by the upper margin), but without any title or indication of authorship. Alternatively, one might assume (as Cribiore probably did) that the passage preserved by the papyrus was an isolated extract copied on a single sheet, yet there is no compelling proof of this, either. Thus, it seems inappropriate to rule out the possibility that the papyrus belonged to a copy in extenso of the Misoumenos, like the others here considered.

Let us turn finally to P.Oxy. LXIV 4407 [15] (Figure 5), which preserves Dis Ex. 1–113 written on the back (upside down) of a register dated AD 241/242. The original roll was trimmed before being reused, as indicated by the fact that the document on the recto is incomplete at the top, whereas both the upper and lower margins of the literary text on the verso survive. Overall, P.Oxy. LXIV 4407 [15] represents a higher quality of product than the other papyri considered here: the number of lines per column (51) seems constant; the margins and the intercolumnar space are quite wide; the line spacing is regular; at the opening of col. iii the indication χοροῦ, centred in the line, is given slightly larger interlinear spacing.

Changes of speaker are indicated by paragraphoi and dicola written in scribendo; notae personarum appear in the left margin of the columns and in the interlinear space. There are ano stigmai, tremata, occasional accents and angular rough breathings. Elision is normally made and marked by diastolai. These lectional signs were probably written currente calamo by the first scribe, as were the corrections and the interlinear additions.^[49] At the left and right sides of the last line of col. ii, respectively, there are the remains of a coronis, marking the end of an act – perhaps the third – and a stichometric note.^[50]

Although the papyrus was already in poor condition at the time of reuse, as clearly shown by the fact that the scribe left a blank space in the middle of the text at the top of col. ii, the presence of stichometry^[51], together with the overall characteristics described briefly here, suggests that this copy of the Dis exapaton was made on commission, similarly to what Lama concluded about other literary texts transcribed on reused rolls.^[52]

Figure 5:

P.Oxy. XLI 2943, Menander, Samia. Courtesy of The Egypt Exploration Society and the Faculty of Classics, University of Oxford.

I will not go into the controversial issue of how official documents could have been accessed for reuse after their expiry date.^[53] But I would point out with regard to the papyri examined here that official documents seem to have been used both for copies of mediocre quality, very probably written privately, as in the case of P.Oxy. XLI 2943 [8] (Figure 6), as well for more careful ones presumably produced on commission by professional scribes, such as P.Oxy. LXIV 4407 [15].

This observation would seem to confirm the opinion of Wilhelm Schubart, who did not rule out the more common hypothesis that the texts written on the verso were private products but who emphasised that many discarded rolls were probably purchased by book dealers to produce copies of lower quality (and lower price).^[54] However, on the basis of our current data we cannot rule out the possibility that private individuals, perhaps those with easier access to administrative archives, could have acquired (or retained) documents, either to make copies themselves, or with the aim of entrusting the writing of such works to professional scribes, not unlike what happened for those rolls reused for other documents (e. g. private letters or accounts).^[55]

Figure 6:

P.Oxy. LXIV 4407. Menander, Dis Exapaton. Courtesy of The Egypt Exploration Society and the Faculty of Classics, University of Oxford.