Richard Janko, University of Michigan
The remarkably detailed story told by Strabo (13. 1. 54) and Plutarch (Sulla 26. 1–2) of the disappearance and rediscovery at Scepsis in the Troad of the library of Aristotle and Theophrastus has strained credulity. Supposedly they were rediscovered by the book-collector Apellicon and taken to Rome by Sulla when he sacked Athens, studied by Cicero in the villa of Faustus Sulla in Campania, worked on by Cicero’s secretary Tyrannio, and catalogued by the Peripatetic philosopher Andronicus of Rhodes in Augustus’ time, when efforts were made to rid the texts of errors that had crept in. Strabo relied on a contemporary witness, Posidonius, who was in Athens when it was sacked, and the testimony of Tyrannio, whom he knew personally. However, his credibility has seemed especially suspect because Athenaeus (1. 3) says that Aristotle’s books were sold by Neleus to the Library of Alexandria and were catalogued by Callimachus.
How could Aristotle’s works have been unknown, even to members of his own philosophical school, for most of the Hellenistic period? The best of the many attempts to solve this riddle (since they best explain all the evidence) have argued that both stories can be right (e.g. Barnes 1997). Aristotle’s works were certainly present in the Library of Alexandria, yet they were at the same time lost, in the sense that his original manuscripts and those of Theophrastus that passed to Neleus of Scepsis lay there forgotten by his ignorant descendants.
Research into the transmission of Aristotle’s Poetics supports this conclusion. There are lacunae in all the witnesses to the text, and errors like that at 1451a33 ἁπλῶν ‘simple’ instead of ἀτελῶν ‘imperfect’, which Essen conjectured in 1878 and survives only as معلولة ma‘lūlati ‘maimed’ in Abū Bišr’s Arabic translation, whereas that used by Avicenna read بسيطbasīṭ ‘simple’. This error was caused by early Hellenistic handwriting in capitals in which TE was confused with Π. Such errors show that the archetype of the Poetics was created in the first half of the third century B.C.E., presumably at Alexandria; it is best reflected in the commentary of Avicenna. However, that archetype was subsequently improved upon by the addition of superior variants like ἀτελῶν, which must have come from Aristotle’s own manuscript. These variants, however, often failed to displace the errors that had spread through most of the textual transmission. It will be argued that, given ancient methods of publication (as witness the publication-history of Cicero’s Academica), this situation exactly reflects what would be expected if the scenario described by Strabo actually happened.