The transmission of ancient wisdom: texts, doxographies, libraries

Ipı: J2v !CUUK925-02 cuuky25Gerson Top: 0.625in Gutter: o875in ISBN: 978 6 521 87642 1 KT mie ee April 2, 2010 1:33 Eee a EVEL > VER 2 OO THE TRANSMISSION OF ANCIENT WISDOM: TEXTS, DOXOGRAPHIES, LIBRARIES GABOR BETEGH ı TEXTS From the authors writing in Greek berween the archaic period and the end of Hellenistic umes, there is only one philosopher whose oeuvre reached us in ies entirety: Plato.’ Although our corpus of Aristotle’s work is far from complete, we have about thirty treatises generally accepted as authentic, and these works contain a significant part of Aristotle’s philosophical output. With the founding fathers of the Hellenistic schools and their immediate followers, we are much worse off. Diogenes Laertius lists forty-one treatises as Epicurus’ ‘best books’, out of which On Nature in itself was apparently almost double the size of the entire Platonic corpus. Of this monumental oeuvre only Epicurus’ brief summaries of his central doctrines have reached us as quoted by Diogenes Laertius. From the early Stoics the only work chat survives is Cleanthes’ Aynın to Zeus, transinitted by Stobaeus. Short fragments and quotations remain of the more than 700 treatises of Chrysippus. No complete work of any of the Presocratics is preserved. The situation is considerably better with Latin texts written around the end of the Hellenistic period with the aim oftransmitting Greek wisdom to the educated Roman audience: we have a number of Cicero’s philosophical works and Lucretius’ poem chat follows closely Epicurus’ On Nature” For the vast majority of thinkers up to the end of the first century BCE ' From che later period, we have Plotinus’ works, due to Porphyry’s editorial activity, and perhaps Marcus Aurelius’. * On top of these texts transmitted through incdicval copies, we have a growing number of fragments discovered on papyri. Most momentous of these are Aristotle’s Arhenaton politeia, found in Egypt in 1890, a fairly long section of Empedocles’ philesophical poem (Martin and Primavesi 1999); and che Epicurean texts found in the library of the Villa of the Papyri in Herculaneum, including at least six of books of Epicurus’ monumental Ov Nature originally in chirty-seven books, treatises by che lace second-contury BoE Demetrius ol Laconia and the first-century Boe Philinteruns, as well as sore fragments of Chrysippus (fer a recent overview ser Sider 2005). Further important papyn are the second-century ct Store [fieracles’ Eleneots of Blues found in Egypt, the anonymous commentaries to on Plato's Theaeretus and Permenites,

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 2 THE TRANSMISSION OF ANCIENT WISDOM: TEXTS, DOXOGRAPHIES, LIBRARIES gabor betegh 1 TEXTS From the authors writing in Greek between the archaic period and the end of Hellenistic times, there is only one philosopher whose oeuvre reached us in its entirety: Plato.1 Although our corpus of Aristotle’s work is far from complete, we have about thirty treatises generally accepted as authentic, and these works contain a significant part of Aristotle’s philosophical output. With the founding fathers of the Hellenistic schools and their immediate followers, we are much worse off. Diogenes Laertius lists forty-one treatises as Epicurus’ ‘best books’, out of which On Nature in itself was apparently almost double the size of the entire Platonic corpus. Of this monumental oeuvre only Epicurus’ brief summaries of his central doctrines have reached us as quoted by Diogenes Laertius. From the early Stoics the only work that survives is Cleanthes’ Hymn to Zeus, transmitted by Stobaeus. Short fragments and quotations remain of the more than 700 treatises of Chrysippus. No complete work of any of the Presocratics is preserved. The situation is considerably better with Latin texts written around the end of the Hellenistic period with the aim of transmitting Greek wisdom to the educated Roman audience: we have a number of Cicero’s philosophical works and Lucretius’ poem that follows closely Epicurus’ On Nature.2 For the vast majority of thinkers up to the end of the first century bce 1 From the later period, we have Plotinus’ works, due to Porphyry’s editorial activity, and perhaps Marcus Aurelius’. 2 On top of these texts transmitted through medieval copies, we have a growing number of fragments discovered on papyri. Most momentous of these are Aristotle’s Athenaion politeia, found in Egypt in 1890, a fairly long section of Empedocles’ philosophical poem (Martin and Primavesi 1999); and the Epicurean texts found in the library of the Villa of the Papyri in Herculaneum, including at least six of books of Epicurus’ monumental On Nature originally in thirty-seven books, treatises by the late second-century bce Demetrius of Laconia and the first-century bce Philodemus, as well as some fragments of Chrysippus (for a recent overview see Sider 2005). Further important papyri are the second-century ce Stoic Hierocles’ Elements of Ethics, found in Egypt, the anonymous commentaries on Plato’s Theaetetus and Parmenides.

P1: JZP CUUK925-02 cuuk925/Gerson 26 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh we are entirely dependent on the indirect evidence of quotations, paraphrases and summaries. What we have today, both in terms of primary texts and ancient secondary sources, corresponds to a considerable extent to what was available in the best libraries at the end of antiquity. Between the end of the first century bce and the third century ce, the situation was significantly different – never after and, more remarkably, never before was the earlier philosophical literature so widely available. The books of most of the Presocratics could be consulted, the works of Hellenistic authors were still copied and taught, new editions of Plato were produced and Aristotle’s treatises had become more easily accessible. This state of affairs was the results of the fact that from the first century bce onwards philosophers showed an unprecedented interest in the texts of authoritative figures of the past which resulted in a sustained effort to organize the oeuvres and to produce canonical editions and commentaries. This activity centred around the works of Plato and Aristotle, but extended also to Presocratic and early Hellenistic authors. This surge of interest in texts was the outcome of an interplay of complex intellectual and institutional developments. One important factor was the demise of the historical schools in Athens, precipitated by the sack of the city by Sulla in 88 bce, and the ensuing decentralization of philosophical life. Although philosophers could organize satellite institutions away from Athens at earlier times as well – as for example Aristotle’s disciple Eudemus did in his native Rhodes – the Athenian schools with their uninterrupted successions of scholarchs functioned as the depositories of tradition and the guarantors of school orthodoxy. Once this institutional setting became defunct, the more or less independent groups and teachers of philosophy around the Mediterranean came to view the texts of the founding fathers of their respective philosophical persuasions (haireseis) as the primary ties to school tradition. The teaching of philosophy was built around the study of authoritative texts and creative philosophical activity started to take the form of exegesis. This stance had important precedents in the Stoics’ attitude towards Zeno, and especially in the way Epicureans treated Epicurus’ writings, but from that time onwards it became ever more prominent among Aristotelians and Platonists. The attitude towards authoritative texts, especially in the Platonic tradition, gradually gained a spiritual dimension: centrally important texts were considered sacred, and their study a religious act. Moreover, there was a growing sense that the classical texts contain the fullest expression of a wisdom that their authors inherited from an even more ancient past. A connected further element was provided by the changing attitude among the Stoics, Platonists and Aristotelians towards the authoritative figures of rival

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 27 schools. On the part of the Stoics there is evidence for a growing acceptance of the authority of Plato and Aristotle which in some cases resulted in the modification of the orthodox Stoic doctrines. The process seems to have started in the late second century bce with Panaetius whom Philodemus calls both ‘philoplaton’ and ‘philoaristoteles’ (Stoic. Hist. col. 61.2–3; cf. Cic. Fin. 4.79; Tusc. 1.79). What is remarkable in the present context is that Panaetius’ admiration for Plato apparently also took the form of a thorough philological study of his texts. Several aspects of this work were known earlier – that Panaetius found different alternative versions of the beginning of the Republic (D.L. 3.37), that he allegedly athetized the Phaedo 63(Asclepius, In metaph. p. 90 Hayduck; Anth. pal. 9.358; Elias, In cat. p. 133 Busse), and that he discussed the particularities of Plato’s orthography (Eustathius, Ad Od. 23.220). Yet a recently rediscovered text of Galen strongly suggests that these were not sporadic remarks, but that Panaetius prepared a critical edition of Plato and treated these, and surely other related questions in conjunction with this edition. Galen’s text also implies that Panaetius’ edition was still available and appreciated for its accuracy in the second century ce.3 In the Platonist tradition the important shift came with the break with the sceptical Academy, and the corresponding desire to present Plato’s philosophy as a closed set of doctrines, on a par with the highly systematized teaching of the Stoics. The justification of this thesis, which soon became the dominant view, created immediate interpretative problems. First, because one had to identify Plato’s doctrines in the dialogues and, second, because different Platonic texts seem to present incompatible views on a wide range of crucially important subjects. Both the determination of the true Platonic doctrines, and the resolution of such apparent inconsistencies required close attention to the relevant passages, including the discussion of the grammatical constructions of sentences and the possible meanings of individual words. In many cases slight textual variations could make a considerable difference. To quote just one example: whether in the sentence at Timaeus 27c4–5 one reads a pair of epsilons or a pair of etas, and if the latter, how those etas are accented, has important bearings on the hotly debated question of whether Plato thought that the cosmos has a temporal beginning.4 In such cases the champions of rival interpretations could defend alternative texts, and thus the identification of Plato’s doctrine 3 Galen, Peri alupias 13, with Gourinat 2008. The early Hellenistic history of Plato’s texts is debated. See e.g., Mansfeld 1994: 198–9. A remark by Antigonus of Carystus shows that around 270 bce the complete oeuvre of Plato was not easily available outside the Academy (D.L. 3.66). It is also worth mentioning that Arcesilaus possessed a copy of Plato’s books (D.L. 4.32; Philod. Acad. hist. col. 19.14–16). 4 Cf. Proclus, In Tim. 1.218.28–219.30 with Dillon 1989.

P1: JZP CUUK925-02 cuuk925/Gerson 28 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh became inseparable from philological questions. In commenting on Timaeus 77b–c, Galen compares different editions of Plato, and singles out the reading of the ‘Atticiana’ – an edition by Atticus, perhaps identical with Cicero’s friend and renowned bibliophile – against the unanimous reading of other editions. What lends special interest to this reference is that the manuscript tradition of Plato’s text known to us has the solitary reading of the ‘Atticiana’.5 People were well aware of the existence of forgeries. Philoponus (In an. pr. 6, 8–10, CAG xiii, 2), for example, reports that ‘[t]hey say that forty books of the Analytica were to be found in the ancient libraries, and only four of them were judged to be by Aristotle’. In establishing the true doctrines of the master, it was thus essential to separate authentic and spurious texts. In some cases, it resulted in the athetization of passages or entire works that we would consider authentic: Panaetius regarded the Phaedo spurious, Andronicus the De interpretatione, and possibly the last chapters of the Categories, while the Stoic Athenodorus of Tarsus, head librarian in Pegamum in the first century bce, expurgated doctrinally problematic passages from the oeuvre of Zeno (D.L. 7.34). We also hear about Epicureans doubting the authenticity of works attributed to the founding fathers of the school (Zeno of Sidon, fr. 25 AngeliColaizzo). On the other hand, later Platonists accepted the Platonic corpus established by Thrasyllus, consisting of 35 dialogues and 13 epistles, arranged in 9 groups of tetralogies (counting the Epistles as one work) – yet the authenticity of a few works included in this canon would be questioned by modern scholars.6 We also occasionally hear about tampering with authoritative texts. Hierocles of Alexandria reports that those Platonists and Aristotelians who objected to the growing tendency to emphasize the doctrinal continuity between Plato and Aristotle had no qualms about tampering with the texts of the founders of their own schools in order to prove more effectively the disagreements (apud Photius, Bibl. cod. 214, 173a; cod. 251, 461a).7 The practice of Athenodorus of Tarsus mentioned above is another case in point. This focus on classical texts resulted also in editions in which a complex system of critical signs flagged textual corrections, suggested transpositions, repetitions, spurious passages, stylistic features, as well as doctrinally important parts and doctrinal agreements (D.L. 3.66; P.Florentina). 5 Galen, In Platonis Timaeum commentarii fragmenta, 3, 2, p. 13, 3–4; with Irigoin 2003: 152. For a detailed but at places dated, discussion of the transmission of Plato’s text, see Pasquali 1952; on the role of the Atticiana, see 278–9. 6 The extent and impact of Thrasyllus’ activity is debated; cf. Tarrant 1993. It is notable that Thrasyllus arranged Democritus’ oeuvre as well. 7 Dillon 1989.

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 29 These changes in the philosophical landscape also provide the background against which we should appreciate the ancient reports about the fate of Aristotle’s and Theophrastus’ texts.8 According to the famous story related by Strabo (13.1.54), Aristotle’s school treatises, together with works by Theophrastus, were hidden in Asia Minor and were not available even for Peripatetic philosophers from Theophrastus’ death until around the beginning of the first century bce, when Apellicon, a wealthy Athenian bibliophile, dabbling in Aristotelian philosophy, brought the whole lot back to Athens. We are told that Apellicon recopied and edited the texts, but apparently did a poor job. The collection then became part of Sulla’s war-booty and was transferred to Rome.9 Sometimes later Tyrannio, the renowned scholar who also helped Cicero rearrange his library, took interest in the contents of the collection – yet it is unclear exactly what he did with it. Plutarch (Sulla 26) adds that Andronicus of Rhodes obtained the material from Tyrannio, and drew up a catalogue of it. Before we come back to Andronicus, let us note that the main upshot of the story for both Strabo and Plutarch is that the eclipse of Aristotelianism in the Hellenistic period is to be explained by the unavailability of fundamental Aristotelian texts. Strabo even adds that those who had access to these texts again for the first time were ‘better philosophers and better Aristotelians’, yet they still could not attain precision in philosophy because their text of Aristotle was defective – no adequate philosophy without adequate texts. It is however notable that Panaetius probably died shortly before the collection reappeared in Athens, so Aristotle could reach the status of authority even outside the Peripatos without the material once hidden in Scepsis; it is probably this shift which raised interest in Aristotelian texts in the first place. Yet, neither Panaetius’ enthusiasm, nor Apellicon, nor the arrival of the collection in Rome, nor even Tyrannio, made Aristotle’s school treatises widely known. As Cicero remarks, few were the philosophers who actually read Aristotle (Top. 1.3). Cicero himself is aware of the school treatises and claims to have consulted them (Fin. 3.3.10), but this – with the probable exception of the Nicomachean Ethics10 – leaves no discernible mark on his presentation of Peripatetic philosophy. The definitive change in this respect was inaugurated when still in the first century bce the Categories started to be discussed across school boundaries. 8 Almost all the details are controversial. For a thorough re-examination of the evidence, with mainly negative conclusions, see Barnes 1999. 9 Sulla’s booty may have contained other libraries as well, of course. We know that his assault of Athens in 86 bce caused some destruction in the Academy, but the school library may well have survived if it had not already been moved to Rome by Philo of Larissa. 10 In the wake of Kenny 1978 here has been a controversy whether the Eudemian or the Nicomachean Ethics was treated as canonical. It seems certain that at least from the time of Aspasius (early second century ce) the Nicomachean (including the common books) was prioritized.

P1: JZP CUUK925-02 cuuk925/Gerson 30 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh Among the earliest interpreters of the Categories Simplicius (In cat. 159.32–3) mentions the Peripatetics Andronicus and Boethus, the Stoic Athenodorus, the Platonist Eudorus, and Ariston of Alexandria, who was a disciple of Antiochus, but then became an Aristotelian (Acad. Hist. col. 35.10–16). From that time onwards the Categories remained at the centre of exegetical literature for centuries, primarily owing to the prominent place it acquired in the Platonist school curriculum. A further sign of the early recognition of the importance of the Categories is that someone in the second half of the first century bce forged a version of it in Doric dialect, and circulated it under the name of the pre-Aristotelian Pythagorean Archytas clearly with the intent to reclaim its doctrinal content for the Pythagorean tradition. The authenticity of this treatise was accepted by all later authorities, except Themistius – authoritative texts could be produced not only rediscovered. But the interest soon extended to Aristotle’s other school treatises – especially in logic, physics and metaphysics – which then soon eclipsed Aristotle’s ‘exoteric’ writings. However, first-hand knowledge of Aristotle remained much less common in later times as well. It is debated, for example, whether Origen had direct knowledge of Aristotle’s texts, and if so how much.11 According to the formerly standard scholarly opinion the breakthrough of interest in the Aristotelian school treatises in the first century bce was due to Andronicus. He was customarily credited not only with producing the first proper edition of the collection acquired by Apellicon, but also arranging books into treatises, rearranging passages, adding bridge sentences and cross-references. Important works – most notably the Metaphysics – were supposed to have received their final form due to Andronicus’ editorial activity, which – it was held – resulted in the authoritative text of Aristotle, standardly used by later philosophers, and forming the direct origin of our corpus Aristotelicum. Crucial elements of this view have been questioned recently.12 Apart from the fact that the relative chronology between the first signs of interest in the Categories and Andronicus’ work is controversial, we have no clear information about either the extent or the exact nature of Andronicus’ activity. What remains certain is that he produced a Pinakes in five books, containing a biography of Aristotle and an annotated catalogue of the oeuvre, which provided a systematic arrangement of the treatises, discussed questions of authenticity, and gave information about their contents. At any rate, Porphyry took Andronicus as one of his models in thematically arranging Plotinus’ treatises in the Enneads (VPlot. 24). Many of the details will remain controversial, but it seems safe to say that later editions of Aristotle’s school treatises were produced on the basis of the material brought 11 Carriker 2003: 85–6. 12 Barnes 1999; Gottschalk 1987.

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 31 to Rome by Sulla,13 and that some treatises completely ignored before – most notably the Categories – started to be discussed with the arrival of this material. On the whole, it appears that there was no single authoritative edition of any of the classical or early Hellenistic philosophers that would have completely supplanted rival editions. Commentators of Plato and Aristotle could compare and discuss the textual variants of different editions, older and newer. This situation can be contrasted with the way in which the Alexandrian editions of poets had rapidly become standard in the third century bce, driving out alternative versions from circulation. Philosophers apparently continued to prepare their own working editions. Galen speaks about the care with which he prepared his own text of Aristotle, Theophrastus, Eudemus, and other early Peripatetics, and his formulation suggests that this practice was not specific to him (Peri alupias 13, cf. Gourinat 2008). Insofar as teaching involved not merely the use of primary texts, but relied also on the commentary literature (cf. e.g. Porphyry, Plot. 14; Alexander, In an. pr. 1.8–9), philosophers and schools needed to build up their own libraries preferably with multiple editions of the most important works. Personal channels could be used to track down good copies for recopying. Cicero’s correspondence gives extensive early evidence for this practice, but Julian will also write to Priscus to seek out all of Iamblichus’ works, because he knows that Priscus’ sister-in-law has a well-corrected copy (Ep. 2, 12.3–5). We have little specific information about the philosophy holdings of the great public libraries in Rome and around the Empire.14 Estimations can be based on the quotations of authors working in specific libraries. Thus, sifting Origen’s and Eusebius’ references may give us an idea of what was available in the library of Caesarea, which Pamphilus built around Origen’s private collection. Such a study can reveal that most of the philosophy books (original works as well as manuals) were from the Roman period, whereas from the earlier literature the library had a fairly good collection of Plato’s dialogues, probably Xenophon’s Memorabilia, possibly some works by Chrysippus, but no Aristotle.15 After the third century, the texts of Presocratic and Hellenistic authors gradually went out of circulation. The same period was pivotal in the later transmission of Greek literature in general, which was determined, once again, by the interplay of intellectual, institutional and material factors. In this process the change of educational curricula in the fourth century, the philological activity in the cultural centres of late antiquity – above all in Athens, Antioch, Alexandria, 13 As Primavesi 2007 argues, the use of letters for the ordering of books – a peculiarity of the Aristotelian corpus – may provide further support for this point. 14 For a conspectus of libraries, see Casson 2001 and Blanck 1992: chs. 8–10. 15 Carriker 2003: ch. 3.

P1: JZP CUUK925-02 cuuk925/Gerson 32 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh Gaza and Constantinople, the not infrequent destruction of libraries in accidental fires and wars – such as the Herulian invasion in 267 ce, imperial cultural politics, the decline of knowledge of Greek in Italy, and the highly complex and fluid Christian attitude towards classical learning were among the determining factors. With respect to the last point, it is worth noting that the views and arguments that Christian authors formulated about the use of pagan literature were mostly at a theoretical level, and had no immediate practical consequences resulting in the loss of particular works of individual authors. What mattered most was that little-read books were not recopied in sufficient numbers and had decreasing chances of survival. From the material side, the change from roll to codex – which became popular among Christians as early as the first century, but became generally adopted by pagan authors only during the third and fourth centuries – meant that texts gradually had to be transferred from one medium to another. Ultimately only those texts survived which were recopied in codex format, but it is hard to assess the specific impact of this process on the loss of Greek texts.16 In the case of philosophical texts the activity of philosophical institutions, the status of an author in the Platonist tradition, and the use of a text in producing commentaries of Plato and Aristotle, appear to be the crucial factors. In the text referred to earlier, Galen says that he also carefully copied ‘most of the works of Chrysippus’ (Peri alupias 13), whereas there is evidence that the Stoic Cornutus about a century earlier inherited the complete oeuvre of Chrysippus from the poet Persius (Vita Persi 5). But with the disappearance of active Stoic philosophers the situation drastically changed.17 As we can see from Epictetus’ remarks, Stoic teaching practice was also organized around the exegesis of the founders of the school (Diss. 3.21.6–7), even though this apparently did not lead to the writing of commentaries.18 The teaching of Stoicism declined by the middle of the third century and we do not hear about practising Stoic philosophers after that time. A century later Themistius informs us that the last available but already damaged books of Zeno, Cleanthes and Chrysippus were recopied in the impressive rescue operation of books initiated by Constantius II in the library of Constantinople (Orat. 4, 59d–60c). But 200 years later, Simplicius can only report that most of the books of the Stoics have already disappeared (Cat. 334.1–3), and what remained surely mainly consisted in the works of the Roman Stoics. Platonism incorporated important elements from Stoicism, but once that was done, it was in no need of early Stoic texts. Epicurean books apparently went out of circulation as early as the fourth century (Julian, Ep. 89b354–5). 16 On the change from roll to codex, see e.g., Reynolds and Wilson 1968; Gamble 1995: ch. 2. 17 Gourinat 2005. 18 Donini 1994: 5090–1.

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 33 Simplicius is also our last informant about the books of the Presocratics. In commenting on Aristotle, Simplicius affixes long quotations to his explanations and shows an unparalleled concern to quote first-hand. This practice makes him our only source for numerous centrally important fragments of Parmenides, Empedocles, Anaxagoras and Diogenes of Apollonia. In order to explain a reference at Aristotle’s Physics 1.4, 187a12–16, and to decide in that context whether Theophrastus or Nicolaus of Damascus and Porphyry are right in the identification of the material principle of Diogenes of Apollonia, Simplicius quotes extensively from Diogenes’ On Nature, providing in the process seven of the twelve known fragments of this author. He notes at the same time, that although he knows of other treatises of Diogenes, this is the only one that he had access to (In phys. 151.20–153.22). A few pages earlier Simplicius quoted more than fifty verses of Parmenides because of the ‘rarity of the work’ (In phys. 144.28). No doubt, such books were extremely scarce. However, without Simplicius’ scrupulousity, we probably would not have guessed that even a single copy of Diogenes or Anaxagoras could still be in existence in the sixth century. As the use of citations and the comparison of different editions by Proclus and Damascius also indicates, the Platonist school in Athens had an impressive collection.19 Yet, the question of Simplicius’ access to rare books is complicated by the fact that he probably wrote the majority of his works, including the Physics commentary, after the Persian exile of the Platonist philosophers. It is unclear where he settled, whether he could still rely on the school’s collection or what other library was available to him. 2 SECONDARY SOURCES Simplicius’ extensive use of original texts is truly exceptional. In the vast majority of cases authors relied on and quoted from secondary sources. These sources, self-standing works or sections in works, may be very different in nature: manuals, anthologies, florilegia of quotations, dialectical presentations and various inventories of philosophical views, that we may collectively call ‘doxographical’.20 Several practical and material factors made the use of such texts highly advantageous, if not inevitable. Finding specific passages or topics in ancient works, especially when they were written on papyrus rolls, was cumbersome. There was no indexing, and references were made only by rough approximations, most 19 On Proclus’ home library, see e.g., Philostratus, Vitae sophistarum 2.21. 20 The Latin equivalent of the term ‘doxographer’ was coined by Hermann Diels more strictly for the authors in the tradition stemming from Theophrastus’ collection of physical opinions.

P1: JZP CUUK925-02 cuuk925/Gerson 34 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh commonly by pointing to a specific book, or books, of a work. Moreover, as the works of an increasing number of authors became more and more difficult to find, these compilations had an ever growing role as the main repositories of information. But the converse is also true: the extensive use of secondary sources must have precipitated the loss of original works. It is always important to keep in mind in assessing sources that different compilations contained shorter or longer citations from original works, so when an author provides a quotation, there is no guarantee that he consulted the original work.21 Of works that could be consulted for information on philosophical views, the most sophisticated ones offered detailed presentations of doctrines together with the arguments supporting them, as well as the historical and dialectical context in which they were formulated. Cicero’s expositions of the ethical (Fin.), epistemological (Acad.), and theological and physical (ND) views of the Hellenistic schools offer a prime example. As we can see from the fragments preserved by Stobaeus, Arius Didymus22 also presented fairly extensive reviews of Stoic, Peripatetic and Academic ethical doctrines, but showed more interest in definitions than in arguments. The presentation of the views of different schools was a popular genre (Peri haireseōn) in Hellenistic times, practised also by authors like Eratosthenes, Hippobotus, Philodemus and Panaetius. Diogenes Laertius (early third century ce?) occasionally also offers relatively detailed summaries of the doctrines of individual philosophers and schools. His work in ten books is a good example of the variety and fluidity of genres and for the way in which information coming from different sources could be combined. Apart from the doxographical sections, the principal stratum of Diogenes’ work is constituted by the biographical tradition. Works in this tradition offered some factual information about a philosopher (provenance, dates, teacher(s), major biographical events etc.), but focused primarily on personal details, anecdotes and memorable sayings that reveal the philosopher’s character, and hence – it was generally assumed – can be just as crucial as his doctrines in evaluating his philosophy. Dates were often based on speculations and the anecdotes made up from the philosopher’s writings. In Diogenes, the proportion of doxographical and biographical material is very uneven in the presentation of individual philosophers. In many cases, he appends further documents, some of which are of prime importance for us: catalogues of the works, letters, wills, poems. And, for a personal touch, Diogenes includes fifty-two epigrams he composed on different philosophers. The work as a whole is structured according to the Successors (Diadokhai) type. 21 For the example of Aristotelian quotations in Hippolytus, see Mansfeld 1992a: 134–52. 22 He may be identical with Arius, Augustus’ Stoic court philosopher; but his date and identity remain controversial. Cf. Hahm 1990 and Göransson 1995.

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 35 Works in this genre – probably starting with the Successors of Philosophers of the early second century bce Sotion – construed the entire history of philosophy on the model of Hellenistic schools. They presented successive generations of philosophers as heads of schools, and then connected the schools in two (or sometimes more) long uninterrupted chains. In the arrangement used by Diogenes, the Ionian line starts with Thales, and the Seven Wise Men, continues with the Milesians, then through Anaxagoras and Archelaus, it reaches Socrates, where it forks into a line linking the Cynics and the Stoics, and another which starts with Plato and then splits into an Academic and a Peripatetic branch. The Italian line begins with Pherecydes, the teacher of Pythagoras, then through Xenophanes, the Eleatics and the Atomists it ends with the Epicureans. Heraclitus was presented as an isolated figure. In general, the establishment of philosophical genealogy emphasizes the importance of tradition. Yet, a lineage in which historical, speculative and interpretative elements are mixed, can reveal more substantive assumptions. That the Academy and the Peripatos are ‘siblings’, and the Stoics are their ‘cousins’, whereas the Epicureans are not part of the family at all, could be widely agreed. On the other hand, the placement of Plato and his successors in one line, and Pythagoras and Parmenides in a separate tradition, evinces a particular stance on Plato that could hardly be accepted by later Platonists. Diogenes does not seem to have a particular agenda apart from presenting everything he can about philosophers. His contemporary, the Christian Hippolytus of Rome, by contrast, provides extensive accounts of philosophers within a highly charged polemical context in his Refutation of All Heresies. Hippolytus’ objective is to prove that the heretics are only echoing the absurd views of the pagan philosophers and, as he explains in the Prologue (1.1.5), he is therefore obliged to expound their doctrines in sufficiently great detail. This rationale makes Hippolytus a valuable source, especially for Empedocles and Heraclitus – yet the wish to emphasize the parallels can distort his presentation in important ways. The texts considered thus far present the material around individuals or schools. An alternative organizing principle is thematic. The compilation of thematically arranged collections of philosophically relevant views had a long history that can be traced back to the fourth-century bce sophist Hippias, whose work Plato and Aristotle also used. Aristotle elaborated the methodology of the creation of such compilations for dialectical purposes (Top. 1.14) and effectively used surveys of available views in his systematic works. Much of the later doxographical material ultimately goes back to Aristotle’s surveys, and to the works composed by his disciples, some of which were specifically aimed at a methodical presentation of earlier views in various fields. Theophrastus’ Peri phusikōn doxōn (it is debated whether it should be translated as The Opinions of

P1: JZP CUUK925-02 cuuk925/Gerson 36 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh Natural Philosophers or Opinions in Natural Philosophy) had a major, although not exclusive role in the subsequent tradition. The most extensive extant instances of this tradition are the Placita (Tenets) of the second-century Ps.-Plutarch, long excerpts in the Anthology of the fifthcentury Stobaeus, and the shorter passages in the Therapy for Diseases of the Greeks of Theodoret, Bishop of Cyrrhus. On the basis of the close parallels among these texts, Hermann Diels has shown in an epoch-making work that they go back to a common source, a work probably written by a certain Aëtius, who in turn was drawing on the anterior tradition, going back to the Peripatetic material.23 The temporal coverage of the Placita literature, just as the Successors, extends back to archaic times – Homer or Thales – to end, for the most part, with the first half of the first century bce.24 Although there is some evidence for an analogous treatment of logical and ethical topics (e.g., Ps.-Galen chs. 9–15), the lists in the Placita tradition deal in a systematic manner with topics that belong to the physical part of philosophy. Ps.-Plutarch, who seems to follow the general structure of his source, covers in 133 chapters divided into five books metaphysical principles, theology, fundamental physical concepts (time, place, motion, necessity and fate etc.), cosmology (shape, generation, destruction, declination of the world), astronomy (e.g., substance, figure, distances, light of the heavenly bodies), meteorological phenomena, geology, psychology (perception, memory, dreams, etc.), physiology, embryology, and some non-human biology. Some of the inventories of views visibly aim at comprehensiveness (for example, the list of material principles in Ps.-Galen ch. 18 has twenty-three items), whereas in other cases it is hard to see why exactly only two or three views are mentioned or recopied. Some of the positions are written out in some detail, but most items are stripped down to skeletal formulations. The views can be arranged systematically or in antithetical pairs (e.g., those who held that the soul is corporeal are opposed to those who held it to be incorporeal), to which compromise or unclassifiable views can be appended. These ordering principles may be combined in the compilation of more complex lists. Those who made use of such collections always did so in composing a work that had its own message, structure and argument. These factors affected in various ways the choice, scope and arrangement of the texts taken over. 23 Diels 1879. Mansfeld and Runia 1997 and 2008 undertake a major re-examination of the evidence. They confirm the fundamentals of Diels’ reconstruction, but provide amendments in important details, such as Diels’ attempt to identify in Theophrastus’ Peri phusikōn doxōn a unique ultimate source of the Placita literature. 24 Sedley 2003: 28; Mansfeld and Runia 1997: 320: the last philosopher mentioned by Aëtius is Posidonius. The same is true for Arius Didymus.

P1: JZP CUUK925-02 cuuk925/Gerson Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 The transmission of ancient wisdom April 2, 2010 37 Authors may copy out just one citation or view that they find particularly apt for their purposes. Or they can copy out longer, uninterrupted stretches, for example a complete survey of the material principles, e.g., Sextus Empiricus reproduces the complete list that we also find in Ps.-Galen at both PH 3.32 and M 9.360–4. Or they can excerpt a whole range of such chapters, as Eusebius in books 14 and 15 of his Praeparatio Evangelica takes over a series of lists from Ps.-Plutarch. If an author found anything to add or update he might introduce excerpts from other texts, either from original sources or, more often, from other handbooks: Stobaeus for instance had a tendency to add quotations from Plato, whereas the Philosophical History of Ps.-Galen is nothing but an epitome of epitomes, partly of Ps.-Plutarch and partly of a different tradition or traditions. Moreover, once someone composed a work in such a way, by drawing on one or more such sources, modifying his source material in whatever way, this newly composed work could now become a source for others. It can also be shown that authors simultaneously used fuller and abridged versions of the same material: Theodoret relied on both Aëtius and the abridged version of Ps.-Plutarch, and Diogenes Laertius used Hermippus both directly and through Sosicrates.25 In this way, a living tissue of texts was constituted, the elements of which were constantly objects of amplification, abridgement, rearrangement and application for a wide range of purposes. On the whole the context-free data of these lists hardly made them apt for a constructive philosophical use. One could certainly quote them as a general display of knowledge or for educational purposes, as Stobaeus excerpts large parts of the Placita, quotes long sections of Arius Didymus, along with selections from poets, historians, orators, to advance the development of his intellectually unpromising son (Photius, Bibl. 112a14–24). Or one could sketch the prehistory of a favoured view: Augustine draws on his doxographical source about the theological doctrines of philosophers in the Ionian tradition (the Successions scheme is at work) to end by saying that ‘it is in order to lead up to Plato that I have summarized these facts’ (De civ. Dei 8.3, trans. Dyson). The arrays of divergent and incompatible views were, however, particularly apposite to advocate the suspension of judgement by the construction of diaphonia arguments. This is Eusebius’ explicit motivation for copying out extensively from Ps.-Plutarch (15.32.9), and in the example mentioned earlier, this is of course Sextus’ reason for presenting the long inventory of material principles – the more formidable the list, the more effective the diaphonia. A final example will bring various topics touched upon in this chapter together. In the De principiis Damascius presents his highly complex metaphysical 25 The methodological problems following from this fact are emphasized in Frede 1999b.

P1: JZP CUUK925-02 cuuk925/Gerson 38 Top: 0.625in Gutter: 0.875in ISBN: 978 0 521 87642 1 April 2, 2010 Gabor Betegh system and discusses the highest levels of reality down to the third member of the intelligible triad. Yet before he continues with the lower levels in his Commentary on Plato’s Parmenides, he turns to show that there is nothing new in the doctrine he has expounded. Not only can one find the expression of the same system in the authoritative texts of the Chaldaean Oracles and the Orphic Rhapsodies, provided one reads them correctly – a topic on which Syrianus had also written – but actually all the ancient ‘theologians’, Greek and Barbarian alike, professed this very same doctrine. In order to demonstrate the point, Damascius avails himself of various sources, but most of his evidence comes from the material that Aristotle’s disciple, Eudemus compiled more than eight hundred years earlier – which in turn relied on the earlier collection of Hippias – about the theogonies of Orpheus, Homer, Hesiod, Acusilaus, Epimenides, Pherecydes, the Babylonians, the Magi and the Sidonians.26 Philodemus used the same material, directly or indirectly, in the first century bce in composing his On Piety, and Cicero again used Philodemus (or Philodemus’ source) for his On the Nature of Gods (1.25–41; cf. Philod. De piet. 3–17). For Philodemus, the Eudemian material ultimately serves the demonstration of Epicurus’ theology. By contrast, in Damascius’ interpretation – which is a late expression of an attitude that can be traced back to Numenius – Eudemus’ collection is evidence for the agreement of archaic sages and thus transmits elements of the same ancient wisdom that can be recovered by an inspired but also philologically attentive reading of Plato’s authoritative text. 26 De principiis 3:162.19ff. Combès and Westerink = 1:319. Ruelle; Betegh 2002.