Since its publication in 1938, the so-called Livre d’écolier has aroused lively interest for two main reasons¹ : first, from a sociological perspective, it is currently the best preserved example of a teacher’s notebook from the Graeco-Roman world. The structure of the book, still clearly visible, enables us to follow a student’s education in Greek from the very beginning with alphabets and syllabaries to the reading and comprehension of literary passages². Then there is its literary content : the Cairo papyrus in fact preserves, within the anthology section, fragments of texts hitherto unknown which are significant for our knowledge of Hellenistic literature : the two adespota epigrams, a comic monologue from an unknown comedy, and a new version of Straton fr. 1³. This paper focuses on different aspects, already touched on by Guérard and Jouguet in their editio princeps but still unresolved, relating to the external structure of the roll and its reconstruction, with constant reference, as a result, to the text’s structure and content.

Let us start with the size of the roll. Presently, only the bottom half is preserved, split into two quite long portions : the left part is 66 cm long, the right 176 cm. The editors calculated the length of the lacuna in between as c. 6,5 cm, making the book’s overall length 248 cm⁴. We know, however, that it was originally longer because the opening portion, unlike the end, does not survive⁵. How much of the roll’s beginning is lost is hard to establish ; nevertheless, we can assume it was at least 22 cm longer in order to display the complete table of consonants linked with each vowel. In fact, 18 columns are missing before the preserved columns of χ and ψ, including the letters from β to φ within the alphabetic order⁶. It is probable, though no textual evidence proves it, that the alphabet was the first exercise of the school book⁷. The original length of P. Cairo JE 65445 would therefore have been ca. 2,80 m, below average for Ptolemaic period rolls, but sufficient to be a school-roll, a type we rarely discover due to the wear and tear which resulted from their constant use⁸.

We can also guess the height of the roll on account of the content of the papyrus, in particular regarding the section containing the passage from Straton’s Phoinikides. Originally, the text occupied at least one column and a half at the end of the anthology and just before the final section containing mathematical exercises⁹. Only the bottom of the two columns survives, with a gap between them of about 11 verses, which we can estimate by comparing the version in Athenaeus¹⁰. We can then assume that the second column was originally 25 lines long and 16 cm high (i.e. the preserved bottom part of 14 lines plus the 11 from Athenaeus). If so, the roll, presumably wrapped, was mutilated roughly in its top half (10 cm) ; its overall height was about 21-22 cm, margins included¹¹. The editors did not attempt to estimate the height of the roll since they did not consider that the exact coincidence of 11 verses between the papyrus’ text by Straton and Athenaeus’ version was a reliable criterion¹² ; this concern is partly understandable when one compares the two texts. Certainly Athenaeus’ text has five more verses spread throughout the first part of the text and missing in the papyrus ; but the second part, on the contrary, does not feature any remarkable textual differences.

A further piece of evidence, however, confirms the roll’s height as 23 cm, for I believe it is also possible to establish the original height of the column which displayed the Macedonian calendar. The previous column ends, interrupted by a table of syllables, with the mention of two Macedonian months : Audanaios and Peritios (19-20)¹³. The editors were not completely convinced of the calendar’s presence in the school text despite the textual evidence, and the lack of available parallels in 1938 compounded this. However, subsequent papyrological discoveries have confirmed that the names of the calendar months occurred in school texts as lists of set words along with, far more frequently, Egyptian calendar names¹⁴. Furthermore, it is notable that in some instances the scribe puts the lists of month names near the syllabaries (simple combinations of two or three letters), as we find it in our papyrus¹⁵. It apparently put the learning of calendars at a very basic stage of ancient education, before memorizing other categories of names (such as deities and geographical or mythological names).

A useful parallel for restoring our calendar section in P. Cairo JE 65445 is P. Louvre inv. N 2328, a contemporary Ptolemaic school text from Memphis¹⁶. It contains a complete list of Macedonian months in the sequence of the agricultural year, starting in September and ending in August (from the month of Dios to Hyperberetaios), as in the Egyptian calendar (from Thoth to Mesore)¹⁷. Since the same sequence is attested also in later papyri which have lists of months, we can conclude that this was the way in which the calendar (both in the Macedonian and in the Egyptian version) was memorized by students at school.

If we accept the height of 23 cm (as discussed above), the lost column after Peritios can be filled precisely with the names of the eight months that follow in the attested list (from Dystros to Hyperberetaios). The bottom part is easily reconstructed since it contains part of a list of 30 numbers, more likely corresponding to the 30 days of the months of the Egyptian calendar – as was argued by Boyaval – rather than to an arithmetical exercise as the editors conjectured¹⁸. The column thus has 21 lines of writing – 8 month-names plus 13 numbers separated by the upper part of the table frame – and is 18 cm high ; together with the margins, it measures ca. 23 cm in height.

The size of the columns, reconstructed using the Straton passage, and apparently confirmed by the calendar section, helps with another difficulty. As we have seen at the beginning, one column is lost in a lacuna between the roll’s two extant parts ; only a few letters from the lower left margin remain. Can anything be conjectured about the content of this lost part ?

The lacuna comes after a completed list of names of Greek deities, and before the end of a list of rivers ; the column must start with a new list of words since the previous column ended with a coronis : after that, a new list of words begins. It is hard to establish the nature of the list as only the first four letters of the last four lines remain legible ; nevertheless, we can assume on the basis of the roll’s estimated height at the top that the list was arranged in ca. 16 lines, i.e. 12 in addition to the 4 extant lines, containing 2-3 words per line¹⁹.

We can guess at the subject matter of this list. First, it is probably a list of proper names which students had to memorize, just as with the other lists on the papyrus ; given this, it seems relevant that it is inserted between a list of deities and a list of rivers. Then, if we compare other school models which contain lists of names, and exclude the categories already present in the text (i.e. Macedonian month names, Greek deities, cities, rivers, and mythological names), we can find some plausible parallels²⁰ : names of Egyptian deities (i.e. following the list of Greek deities) or more likely other toponyms (i.e. islands, mountains, but perhaps also springs, sanctuaries and sites more connected with religion)²¹.

The end of this unknown list is clearly marked by a coronis : what follows (another six lines of text at 52-57) is, according to the editors, the beginning of the list of rivers set in the next column, since in their reconstruction they integrated the few letters which are preserved with the names of rivers and the π in eisthesis with the title Ποταμοί (52)²². This conjecture is questionable. First, it generates the repetition of the name Сτρυμών at 57, which occurs also at the end of the list in the second column (66). Such a mistake is not implausible (even if it should be taken as extrema ratio in the interpretation of a rough passage), but the editors supposed the presence of at least two such additional mistakes in the same list of rivers²³. It seems unlikely that a scribe, who has shown himself impressively competent in the rest of the text, should make one repetition and two misspellings in the space of ten lines. Then the list of rivers, as the editors reconstruct it, appears excessively long. If we consider the fact that the two columns have three words per line each, and that at least ten lines are lost in between, the whole list would have contained roughly 65 rivers²⁴. This seems like too much, compared with the other lists in the papyrus and in view of the function of such a list, namely to memorize and learn relevant names.

Alternatively, the end of this column could contain the beginning of a list of cities, in which case the list of rivers was originally limited to the following column. Let us start with the textual evidence. The last three lines can be restored with three names of cities²⁵ : Ἴсμαροс (55), Ἐρύμναι (56) and Сτρύμη (57) ; the extant Π of the title would be part of πόλειс²⁶. All three toponyms are attested in literary sources and well-known in lexicography²⁷. Moreover, Stryme and Ismaros are in Thrace and Erymnai in Lycia ; checking the following list, almost all the rivers mentioned (13 out of the extant 15) are located in the same areas : Thrace/northern Greece and the Anatolian peninsula²⁸. It is unclear whether this indicates a marked interest in particular regions, or simply reflects the compiler’s use of a geographical sourcebook concerning these regions²⁹.

Indeed, this reconstruction, which excludes any mention of the further parts of Asia, challenges the idea that the Livre d’écolier is the expression of the geographical knowledge typical of the early Ptolemaic era, influenced by, if not based on, the conquests of Alexander the Great³⁰. The primary purpose of the toponym list was presumably to help a pupil to read literary works by memorizing unusual names common in Greek literature, from Homer to learned Alexandrian poetry³¹. This perspective sets our papyrus apart from the genre of paradoxography which the other almost contemporary Livre d’écolier, known as the Laterculi Alexandrini, represents³². This may then reflect the existence of different strategies in the education system of the early Ptolemaic period for teaching pupils how to master the classics of Greek literature.