Revising doxography: Hermann Diels and his critics

ET Js 2001 | 2 | 219-283 L_ 2335 ZAMS, L- Maa, : LEONID ZHMUD gnuen REVISING DOXOGRAPITY: HERMANN DIELS AND HIS CRITICS* Our knowledge of the Presocratics, except for their verbatim fragments, is mainly based on the two “pillars”, Aristotle and Theophrastus, whose reliability as sources is crucial for our understanding of early Greek philosophy. An attempt to offer a substantially new interpretation of an individual Presocratic or of Presocratic philosophy as a whole almost inevitably involves criticism and reassessment of these sources. Recent books by A. Capizzi and P. Kingsley are characteristic examples of such an approach!, Both authors adopt the assumption that Aristotle not only interpreted the Presocratics in his own terms and for his own goals, thereby transforming their views, but more significantly that he could not adequately understand the Presocratic “mystical” mentality. I do not sec, however, any serious grounds to believe that the worlds of the Presocratics and of Aristotle were incompatible. It is probably not by chance that the most systematic and well-founded attempt to reassess Aristotle as a source of Presocratic philosophy, that of Cherniss?, was not tied to any specific reinterpretation of the Presocratics themselves. Later J. McDiarmid tried to apply to Theophrastus’ doxography the principles worked out by Cherniss3. He stressed once more the dependence of Theophrastus’ thought on that of Aristotle, but exaggerated its consequences for reconstructing Presocratic philosophy. Along with Aristotle and Theophrastus, Hermann Diels can be viewed as another, more immediate, and no less important source of our knowledge about the Presocratics. By his brilliant restoration of the doxographical tradition in the Doxographi Graeci (DG) Diels laid a firm foundation for his later edition of the Fragmente der Vorsokratiker, on which the whole of 20%-century scholarship in this field rests. Due to the fact that after Dicls astonishingly little was done to continue his research in the doxographical sources, his authority lasted unchallenged for almost a century, though * Oral versions of this paper were presented at Lille University (October 2000) and at St. Andrews (April 2001). | am very grateful to all who participated in the discussions. ' A, Capizzi, The cosmic republic, Amsterdam 1990 (La republica cosmica, Roma 1982); P. Kingsley, Ancient philosophy, mystery, and magic: Empedocles and the Pythagorean tradition, Oxford 1995. ? HL Cherniss, Aristotle's criticism of Presocratic philosophy, Baltimore 1935, * 4. B. McDiarmid, Theophrastus on Presocratie causes, HSCPh 61 (1953), 85-156.

L.. Zumup, Revising Doxograpby: Hermann Diels and his Critics P. Steinmetz.” attempt to question an important point of Diels’ reconstruction can be retrospectively viewed as a sign of the approaching revisionism’, In 1984 A. Lebedev proclaimed a full-scale offensive against Diels’ doxographic al theory: 221 What are the fundamental Dielsean assumptions under attack, as stated in a recent paper by Runia!% First, there is the notion of a doxographical genre. Similarities : between the doxographical texts, Runia believes, are better explained not in terms : of literary genre, but in terms of philosophical method, i.e. Aristotelian dialectic. i Second, “Diels showed insufficient awareness of the possible tension between the Diels’ bistory of doxography is false throughout. “Aëtius” is a phantom ..., and Theophrastus as a pandoxographic bypokeimenen is à mastodon of the XIX-century Emquellentbeorie. À e omplete revision of Diels’ theory is in preparation. The Placita philosophorum have nothing to do with Theophrastus, but are of Academic (Eudorus) and Stoic (Arius Didymus) origin. We know almost nothing about Theophrastus’ obscure work Puoueai SEA’. + methods of ancient doxographical texts and the requirements of the modern study of [the history of philosophy”. We can only exploit this material “when we have understood the aims and method of these texts”. Third, Diels was too critical of the later | writers, accusing them of making all manner of unnecessary changes: “he failed to recognize the flexibility of a doxographical method”. Closer reading of the works by Mansfeld, Runia and Baltussen for the last 15 years Later Lebedev changed his mind and never completed his promised “complete revision”, but his Russian translation of the first volume of the Vorsokratiker bears reveals that there is much more under attack: not only Diels’ understanding of the some traces of his deep disagreement with Dielsé, In a recent book by the leading condoxographical tradition, but also his integrity of as a scholar and even the relevance of temporary experts on doxography, J. Mansfeld and D. Runia, I found an absolutely German classical philology of the 19" century!!. I shall try to enumerate what exdevastating critique of Lebedev's ideas’. Equally unsuccessful was the other interactly the new doxography does not like in Diels, taking into account mostly those mediary critique of Diels, that by Kingsley, whose “unfortunate and totally unsubstantial remarks have been refuted by Mansfeld in a critical note”*. Although Mansfeld and Runia have repelled some, in their view, inauspicious attacks on Diels, their general approach to his work is highly critical, and hardly any part of Diels’ reproaches, which are made repeatedly by all three scholars. To begin with, the term “doxography” is flawed. Then, Diels’ understanding of doxography is wrong (since it is not a genre), as well as insufficient and too narrow, whereas the modern usage of the term is too wide — not least because of Diels’ lack of clarity. His Prolegomena to the doxographical theory is left unchallenged in their numerous works. Mansfeld, Runia DG received a standard epithet: it is “labyrinthine”, whereas his reconstruction of the and later H. Baltussen® initiated a revival of doxographical studies and created a doxographical tradition is “imposing” but “quite problematic”. Diels “slavishly follodistinctive school, which in view of its essential differences with and conscious distancing from the “old” doxography could be called “a new doxography school”. wed” his teacher Usener, while his references to his other predecessors are both “par- This paper will concentrate mostly on points of disagreement between Diels and his one hand, to Aristotle as a founder of doxography, especially to his dialectical simonious and obfuscating”. Furthermore, Diels devoted too little attention, on the critics, leaving aside for the moment all their important and valuable contributions to method, and to the methods and aims of the late doxographers, on the other. At the doxographical studies and to classical scholarship in general. same time, he paid too much attention to the now doubtful method of Quellenforschung and especially to the obsolete Einguellentheorie. His understanding of the title of Theophrastus’ doxographical work is wrong, as well as his ideas about its origin and structure. Contrary to Diels, Theophrastus’ De sensibus was not a part of the PvoikOv 0084, whereas most of the fragments attributed to the latter work come ' P. Steinmetz, Die Physik des Theophrastos von Eresos, Berlin 1964, 334-5| (sec below, p. 231). ° A, Lebedev, bia rekuvretovoe: Neglected fragments of Democritus and Metrodorus of Chios, in: Proc. of the Ist intern. congress on Democritus. Xanthi 1984, 14 and n. 1. " Egs he gave a different title to his translation, “The fragments of the early Greek philosophers”, and included among them Abaris and Pythagoras’ wife Theano (both are not historical fipures). from Theophrastus’ Physics. Of course, the “youthful Diels did do a splendid job”, but he was just “a son of his time”, so it is odd that his “methodologically flawed” reconstructions, which were never properly criticised, still dominate modern scholarship 1?. * J. Mansfeld, D. Runsa, Aétiana: ‘The method and intellectual context of a doxographer, vol. 1, The sources, Leiden 1997, 333-38 (further Aëtiana). * Aëtiana, 154 n. 109, CF, P. Kingsley, Empedocles and his interpreters: The four-element dexography, Phronesis 39 (1994), 235-54. Another critical note, J. Mansfeld, Doxographi Graeci, in: W. M. Calder ni, J- Mansteld (eds.), Hermann Diels (1848-1922) et Ja science de l'antiquité, Vandeeuvres; Genève 1998, 164, is devoted to the thesis that Aëtius could not bear this name, see: J. Bremmer, Aétius, Arius Didymus and the 1 1). Runia, What is doxography?, in: P. van der Eijk (ed.), Ancient histories of medicine, Leiden 1999, 38-39: “Fundamental assumptions under attack”. Wo“ the theoretical foundations of Diels’ imposing edifice firmly belong within the German intellectual transmission of doxegraphy, Mnemosyne 51 (1998), 154-60. climate of the 19 century” J. Mansfeld, Studies in the historiography of Greek philosophy, Assen 1990, 23). * HH. Baltussen, Theophrastus on theories of perception: Argument and purpose in the De sensibus (Diss.), Utrecht 1993 (further Theophr. 1). Cf. a revised version of this book: Theophrastus against the Presocratics and Plato: Peripatetic dialeetic in the De sensibus, Leiden 2000 (further Theophr. I). riography of Greck philosophy. This is odd, since the Doxographi Graeci is an outstanding implementation of 12 Mansfeld (n. 11), 22: “Diels’ reconstruction ... still maintains its position as a commonplace of the histothe method of Quellenforschung, which method today does not enjoy a good reputation”. Cf below, n. 77.

This list is far from exhaustive, but I need to stop somewhere to have an opportunity for at least a partial reply to this criticism, It is true that in the last hundred years almost everything written by Diels on doxography was considered correct, if the opposite case has not been proven®. If this entails “Diels’ dominance”, this only reflects the real situation: he has done more for the study of the Presocratics in general and of doxography in particular than anybody else. It is therefore very regrettable that the struggle against Diels’ dominance, which became a driving force of the whole offensive", is completely changing the approach to his legacy. Too much of his work, which has not been replaced by anyone else, is taken for granted, whereas every minor correction of his position and every step beyond it is celebrated as a triumph of Philologus 145 (2001) 2 223 " need to argue that such a theory is impossible, and luckily for us, otherwise its : dominance would be truly unbearable. This docs not mean that I plead for a conf venient relativistic approach and propose to judge Diels’ theories only by the standards of 1879, when the DG was published. On the contrary, it turns out that i most of what he did is still valid and that very often his results are more convincing A than those of his critics. The term “doxography”, devised by Diels, is criticised from various points of view. È Mansfeld and Runia are inexhaustibly inventive in adducing ever new arguments that | replace their old ones, in order to show why it is so bad. These arguments include, E e.g., the following: a modern scholarship. In a lecture given at Leiden University during a symposium held to celebrate the publication of the first volume of Aétiana, D. O’Brien compared the field of doxography (which Diels tried to cultivate so hard) to a minefield, but then reassured all the students: But have no fear. Jaap Mansteld and David Runia are on their way. Let us wish them well in their brave resolve to render that dangerous terrain a sater place for the innocent and unwary ®. I still have my doubts. Specifically, I am afraid that this critical campaign that is obviously gaining momentum will become a flourishing business with such self-destructive results that even Jaap Mansfeld and David Runia will hardly be able to compensate for them. I] In spite of the repeated usage of the word “methodology”, an ideal of a scholarly theory which lies behind the attacks on Diels is more then just methodologically flawed. It is a holistic ideal of a theory that penetrates its subject from top to bottom and exhaustively explains it away'® An author of such a theory has to predict all the possible changes in scholarly interests in the next 150 years in order not to be accused of failing to exploit some idea or not paying enough attention to some fact. I hardly 4 It is not true, however, that he was never properly criticised. Even Dicls’ devoted student, O. Regenbogen was not blinded by loyalty to his teacher. His RE article contains numerous criticisms and doubts The term doxography ... is rather unfortunate because, if only by association, it suggests something as solid, impersonal, and objective as, say, cartography", Following this associative criticism one can reach such an absolutely unsolid term as “pornography”, which in itself hardly compromises Diels’ neologism. What follows then from the repeated stress that it isa neologism ? Such an objection would permit us to study Greck culture only with its own terms, which is an impossible task. In the title of G. Lloyd’s book, “Science, folklore and ideology”, we have three neologisms; that reflects the fact that although the ancient Greeks did have science, folklore and ideology, they did not have the general terms for them appropriate for our analysis of these phenomena. In fact, “doxography” has proven its right to exist and has every chance of winning a competition with the alternative neologism coined by Baltussen, “critical endoxography”!8, which has not been used even by Mansfeld and Runia. Mansfeld and Runia criticise the excessively broad usage of the word “doxography” since “its meaning has become quite diffuse”. They even claim that because of this “not a few scholars prefer to avoid the term (and the subject) altogether” '’. | think they overdramatize the situation. Broad usage of a word does not necessarily lead to confusion. There are hundreds of scholarly terms with many clearly recognisable meanings, so that a new meaning, such as “doxography” applied to the modern exposition of the ancient theories, can peacefully coexist with the original one without obfuscating it. Anyway, Diels obviously does not bear responsibility for the broad understanding of the term “doxography”, since he understood it quite narrowly and seldom used it2. Nevertheless, Mansfeld and Runia note: “To be sure, Diels himself about Diels’ doxographical theory: Theophrastos, RE Supplbd. 7 (1940), 1535-38. 14 “I is a fairly recent phenomenon that the study of ancient historiography of philosophy has started hberating itself from the dominating influence of Diels. In mare recent years the flaws of this reconstruction and the Einquellentheorie as a whole have come under attack” (Baltussen, ‘Theophr. 1, 7). Cf. the chapter 1? Mansfeld (n. 11), VIIL “Diels’ error” J. Mansfeld, Heresiography in context. Hippolytus’ Elenchos as a source for Greek 1# Baltussen, Theophr. 1, 21; id., A ‘dialectical’ argument in De Anıma A 4, in: K. Algra et al. (cds.), Polyphilosophy, Leiden 1992) and the paper: H. Baltussen, Plato in the Placita (Aétius Bk. IV): A Dielsian blind histor. Studies in the history and historiography of ancient philosophy presented to J. Mansfeld, Leiden 1996, spot, Philologus 144 (2000), 227-38. See also below, n. 15. 335 n. 14. Cf. ibid., 333 n. 6. 15 D. O'Brien, Hermann Diels on the Presocratics: Empedocles’ double destruction of the cosmos (Aétius 11 4.8), Phronesis 45.1 (2000), 16. '* “Diels’ treatment is incomplete and so inadequate” (Baltussen, Theophr. 1, XXI). CH. below, n. 94. 19 Aétiana, XIII. 22 The term “doxographisch” appears for the first time in 1893, 14 years after the DG: W. Burkert (ed.), If. Diels, Kleine Schriften zur Geschichte der antiken Philosophie, Hildesheim 1969, 101.

L. Zumun, Revising Doxography: Hermann Diels and his Critics Phil gus 145 (2001) 2 225 the existence of a biographical genre. What, in this case, are the reasons for connecting contributed to the confusion”*!, Strangely enough, they find this contribution in the only footnote to Diels’ Ælementum2, Il remains unexplained how a single word in a footnote in the seldom consulted book that docs not deal with doxography could | discuss Atadoyat in the work entitled “What is doxography?”-5, since all scholars have misled anyone. What can be really misleading is not a diffuse usage of the word “doxography”, but ' working on Successions-literature unanimously placed it in biography?”. Indeed, any generic classification has to take into account the origin of the genera, rather a rejection of doxography asa distinctive genre, which in turn leads to an arbitrary categorisation of diverse types of writings as doxographical. Indeed, since Mansteld and Runia rejected Diels’ “rigorous distinction” between doxography and biography’, their concept of doxography became much broader than that of Diels, sometimes it is even hard to grasp how they define it. Diels’ view on doxography might appear too narrow, but at least it was clear: doxographers are those writers who compiled books the content of which could be traced back to Theophrastus’ Bvorkiv dOEaL. Diels believed in the existence of the doxographical genre, which is represented mainly by the Placita-literature, but is not confined to it, as no one did or would call Theophrastus’ work GoéoKovtu or placita. As a matter of fact, I could not find in the modern criticism any serious arguments that Diels was wrong and that this genre did not exist. What can be found amounts to claiming that we prefer to analyse the material not in terms of a literary genre, but in terms of a philosophical method, which I shall discuss later. There is no doubt that literary and folklore genres exist objectively, even if their and not only similarities between them. In the case of the Successions-literature, ; Runia seems to ignore the last criterion, for doxai are presented here even to a lesser degree than in such non-doxographical genres as commentaries, encyclopacdias, ; introductions, ete. In the case of the literature On sects, Runia does not pay sufficient attention to the first criterion, viz. to the question whether this type has any genetic kinship with the doxographical genre originated by Theophrastus. Hippobotus was the author of two books, one biographical, “Avayougy ta gràoodgiv, and the other historiographic, Nepi aipéotev, devoted to ethical doctrines of the Postsocratic schools, If we bear in mind that Theophrastus’ work was concerned with the physical doctrines of the Presocratics and Plato, this rules out regarding Hippobotus’ On sects as a successor to the votkdv d6Eat. In which sense then is this work doxographical? The question can be generalised: According to what methodological borders are not so strict as it seems in theory, and there are many intermediary or mixed works. Relying on the obvious differences between doxography and biography, Diels did not include in doxographical genre either the Successions-literature (Atadoxat), or the so-called Ilepi aivéoewv literature. He placed both Sotion, the originator of the Succession-literature, and Hippobotus, the author of the earliest known work Ilegi aigéoewv, among the biographers’, in full accordance with what we know from the fragments of their works?6. Now, even if Mansfeld and Runia do not believe in doxography as a genre, they would probably not go so far as denying with doxography typically biographical Atadoyai?’? | do not see any reason even to criteria does Runia distinguish doxography from the other kinds of writings pertaining to the historiography of philosophy? To be sure, he is not the first to relate the literature On sects to doxography?!, but in view of his warnings against diffuse usage of this term, one would expect a more conscious approach to the problem. 7 In his DNP article Runia treats three types of literature: Placita, On Sects, and Audoyat. The latter is introduced as follows: “Eine dritte Art von Schriften, die man mit der Doxographie verbinden kann, ist die sog. ‘Sukzessionen’-Lit.”; cf.: “Diese Art zu Schreiben ist zwischen Biographie und Doxographie angesiedelt” (D. Runia, Doxographie, DNP 3 (1997), 804-05). 2 Here Runia examines the same three types as in his DNP article, which gives an impression that all of them are doxographical, but then concludes: “Accounts of successions of philosophers which do not focus on doctrinal differences and developments also fall outside the scope of doxography” (52), which again leaves to 2 Adriana, XII the reader to decide whether there are such successions that do focus on doctrinal differences and develop- 22-14. Diels, Elementum: Eine Vorarbeit zum griechischen und lateinischen ‘Thesaurus, Leipzig 1899, VII ments and theretore belong to doxography. * W. von Kienle, Die Berichte über die Sukzessionen der Philosophen in der hellenistischen und spätn. 2. He calls “doxographisch” the first part of the book Verer:in de elementis placita. 2° Mansfeld (n. 11), 226 “The term ‘doxography’ ... was invented by Hermann Diels ... as the designation of what he believed to be a specific genre of ancient writings” (Aëtiana, X111). “The unspoken assumption of an identifiable doxographical genre that is broader than merely the Placita-literature is one of the chief reasons for the unclarity surrounding the notion el ‘ancient doxography”” (Runia (n. 10), 36). “The genre of doxoantiken Literatur (Diss.), Berlin 1961, 79 ff. J. Meier, Diogenes Laertius and his Hellenistic background, Wiesbaden 1978, 64; Wehrli (n. 26), 13; Giannattasio Andria (n. 26), 18. 2 Mejer (n. 29), 77. Succossions literature and that On sects). There is some truth to this distinction, but generally it does nor hold” 4 B, Wyss, Doxographic, RLAC 4 (1959), 199, Wyss considered it a modified form af ‘Theophrastus’ doxography, whose main representative was Arius Didymus. The problem with Arius is, however, that as the recent thorough analysis shows he wrote two books: Ilepi cipéormy, arranged by topics, and “Europ, (J. Mansteld, Sources, in: A. A, Long (ed.), The Cambridge companion to early Greck philosophy, Cambridge organised by schools: D. E. Hahn, The ethical doxographies of Arius Didymus, ANRW 11 36.4 (1990), 1999, 35). 2 Runia(n, 10), 38-39. 3012ff., 3018 ff; cf. D. Runia, Arcios Didymos, DNP 1 (1996), 1041-42. Meanwhile, it is che arrangement by schools that is universally regarded as a specific feature of the literature On sects! Cf. “the designation as doxo- 2 Diels, DG, 147 f., 245; “Hippobotus biographus” (ibid., Index, 684). graphy of the remains of the literature belonging to the genre On Sects dealing with cthies is a sort of solecism, 2 E Wehrli (ed.), Sotion, Die Schule des Aristoteles, Supplbd. II, Basel 1978; M. Gigante, Frammenti di Ippoboto, in: Omaggio a Piero Treves, Padua 1984, 151-93; R. Giannattasio Andria, | frammenti delle “Sucthough probably an incradicable one” (J. Mansfeld, Doxographical studies, Quellenforschung, tabular presentation and other varieties of comparativism, in: W. Burkert ct al. (eds.), Fragmentsammlungen philosophicessioni dei filosofi”, Napoli 1989, scher Texte der Antike. Göttingen 1998, 26), graphy, in Diels’ view, was to be sharply distinguished from fanciful biography (in which he included the

Unlike Runia, Mejer in his analysis of the Hellenistic historiography of philosophy does not include the works Heoi dipformv in doxography, but treats them, along 11 226 with the Successions-literature and biography, as separate types of writings. In this he closely follows Ueberweg-Pracchter’s Grundriss which distinguished the same four groups in the ancient history of philosophy: 1) biography; 2) Successions-litera- | ho h NI What was the title of Theophrastus’ collection: Pvotkmv dDÓZOAL or Buorai det? This is after all a minor point, but it is intrinsically connected with the more important one, whether Theophrastus’ approach to the Presocratic doxai was more histoture; 3) doxography; 4) Ileoi aipfoewv literature’, There is no wonder that Praechter, rical or more systematic. In late doxography, say Mansfeld and Runia, the doxai are in his turn, heavily relies on Diels’ DG and highly praises his work. What is really suroften more important than names, and this makes them believe that such was Theoprising is that these four groups are to be found in Mansteld’s 1999 work (without phrastus’ attitude, too, which had to be reflected in the title of his book, Duoikui reference to Ucberweg-Praechter). Even more puzzling is that Mansfeld unexpected- : DOEuL. To deduce the decisive argument from the ude as it is given in the list of his ly regards them as genres, + writings (Pudo doo ız’) is impossible, but there is strong evidence in favour of Pvarkov dOEGL as much more historical and person-oriented work than is suggested as if his previous criticism suddenly evaporated: We should therefore look a bit more closely at the various ancient genres which, in a loose sense of the word, we may call historiographic ...: (i) doxography, (ii) biography, (ni) literature on scets (Pert Haireseön), (iv) literature on the successions of the philosophers ... (Diadochas) ...* To the four traditional historiographic genres Mansfeld adds two new ones: “(v) collections of maxims (gromai), apophthegms, anecdotes, pronouncement stories (chreiai) ..., and (vi) introductions (eisagögat)", but then admits: “collections of maxims and anecdotes ... are not a historiographic genre in the proper sense of the word”35, The same can be said, indeed, on the Eisagogai, so it is only natural that Mansfeld further mentions these two genres only in passing. It seems hardly possible that Mansfeld was unaware of the conclusion reached by Runia in the same year: “There was no specific genre of doxographical writings in by the critics)”. Usener and Diels believed the titles Puctkwyv d0Ewv in 16 books and Ilepi quorkov in 18 books to be the same work*. Now, Fepi guoLk®v is not about ta prorká, but about oi quatkoi, who represented for Aristotle, his students, and his commentators a distinctive group of thinkers, whose opinions were the subject of doxography*!. Later authors quote this book under different titles, but all the quotations concern opinions of the specific quoukoi, and not just anonymous “physical tenets”#2, Whereas in Aristotle we find some examples of @uatkdav d0€018, quoixai dota are not to be found. Mansfeld’s favourite and sole example, guatkai TOOTAGELG (Top. 105 b 20ff.), can not be considered as a substitute for it, because the division of the dadektIKai AOOTÚCEL into cthical, physical and logical according to the subject does not confirm existence of the alleged class of guoıkal d6Ea#. Aristotle’s comantiquity, and Diels was wrong to give that impression” 36. In any case, Runia in this very text encourages the reader “to consult two state-of-the-art survey articles by J. Mansfeld” 37, one of which I have just quoted. It turns out, hence, that the state of the art in 1999 has not progressed much from 1920, indeed, it became much more confusing. | would not press too much on unanimity, if Mansfeld and Runia had not announced in 1997 that they “have reached full agreement on all the main issues, and “ It seems that they do not consider their arguments as decisive: “It is however doubttul whether ... the Theophrastean title should run “bvoexdv’ rather than ‘®uovcai dúZal” (J. Mansfeld, Doxography and diahave agreed to differ only on one or two minor points”, But what issue can be more lectic: The Sitz on Leben of the ‘Placita’, ANRW 11 36.4 (1990), 3057 n. 1); cf. id. (n. 23), 23. “As Mansfeld has pertinent for any work on doxographical tradition than to define, what is doxoshown ..., the title is more likely to have been Pvorxai bE” (D. Runia, Xenophanes or Theophrastus? An graphy? Aétian doxographicum on the sun, in: W. Fortenbaugh, D. Gutas (eds.), Theophrastus of Eresus: His psychological, doxographical and scientific writings, New Brunswick 1992, 116 n.9); cf. id. (n. 10), 38 n. 24. 4 H. Usener, Analecta Theophrastea (1858), in: Kleine Schriften, Leipzig 1912, 27; Diels, DG, 102; Regen- 2 Mejer (n. 29), 75 ff. © Fr. Ucherweg, K. Praechter, Grundriss der Geschichte der Philosophie, 1. Teil, Die Philosophie des Altertums, 11. Aufl., Berlin 1920, 17 1., 20 ff. M J. Mansteld, Sources, in: K. Algra et al. (eds.), The Cambridge history of Hellenistic philosophy, Cambogen (n. 13), 1395. # Cf. another Theophrastus’ title, Huds tots quvaikots (D. L. V, 46). 42 Alexander’s variant, Dept tev grate (fr. 227C EHSG), can only mean On the Physicists. The subject of this book are S6ERL tv yuoLKOY: Real Meapperidor Kal Tig DON aùToù Kai Oedq¢yaotos ... Akycı. Galen bridge 1999, 16-17, Mansteld’s further note, “Le should however be borne in mind that these genres are not gives the correct title and refers to Thy Zevog avovg dótar (fr. 231 FHSG). Taurus’ variant, Heot ré prov rigidly distince”, corresponds to the note in Ucberweg-Pracchter: „Innerhalb der ... philosophiegeschicht- Boca (fr. 241A FEISG), can mean both On the Physical Tenets and On the Opinions of the Physicists. In view lichen Literatur kann man im allgemeinen vier Gruppen unterscheiden, ohne daß sich zwischen diesen überall of the other evidence the latter variant is preferable. Cf. J. Mansfeld, Physikar doxai and Problemata physica durchaus scharle Grenzen ziehen ließen“ (20). Irom Aristotle to Aétius (and beyond), in: Fortenbaugh, Gutas (n. 39), 64-65; R. W. Sharples, Theophrastus of 35 Mansfeld (n. 34), 20. Eresus. Commentary, vol. 3.1, Sources on physics, Leiden 1998, 10. + Rania (n. 10), 52. 13 > Runia(n. 10), 33 n. I. * Arıstotle calls this division “tentative”, and it is never repeated in his writings. What he defines here is the 8 Actıana, XXI. own DOC Tv gora (Phys. 187 a 28), Y TOv guearoA bye dbza (Met. 1062 b 22). subject of the dialectical propositions and not the subject of all the possible d6Etu, pace Mansfeld (n. 23), 29.

mentators do not seem to know of such class too: there is a lot of quo OOEUCL) in 229 Theophrastus’ work was based, at least partly, on his critical monographs devoted + to the different Presocratics - Anaximander, Anaximenes, Anaxagoras, Empedocles, CAG*, but no qiuourai dela, Mansfeld and Runia have not brought any convincing argument against the tradi- Archelaus, Diogenes, Democritus, Metrodorus (D. L. V 42-44, 49)°%, Why would he tional view that the De sensibus (DS) originally belonged to the buouxay dÓEM, à have written so much against or about individual quouot (cf. above, n. 41), if “the while Baltussen, after long discussion and not without inner struggle concludes: “If i tenets are what matters most, not the persons who held them” 4? One can assume that pressed one would be tempted to attribute the DS to the Physikai doxai (as characterised in recent research)”"#, Now, the DS was obviously more person-oriented and historical than late doxography*. Whereas sections in Aétius contained different doxai on each sense, Theophrastus arranged material by theories of the individual thinkers on all the senses, which he discussed and criticised at some length. At the very beginning Quowoi are divided into two groups, according to their dominant principle. The order of the names of those who followed the principle “like by like” is “when reorganising this material for the Pvaucdy O6EuL, Theophrastus did not consider it necessary to abandon his critique, although he probably reduced it. Judging E by the extensive criticism in the DS, he was not primarily concerned with the logical b consistency of a theory, although he did pay attention to this point. His favourite ' argument was rather that a theory contradicts the facts (observations) or does not explain the obvious facts, In other words, his criticism was scientific, and not dialectical, as far as a scientific theory has be consistent both with the known facts and Parmenides, Plato, Empedocles, which reflects the growing complexity of their theories®, Yet they are followed not by the proponents of the opposite principles, but by with logical rules. all the others, arranged in the chronological order: Alemacon, Anaxagoras, Clidemus, order, gives their patronymic, birthplace, and the name of their teacher and indicates Diogenes, Democritus*!. The same combination of the topic- and person-oriented approach is characteristic for Menon’s "lartpırn) ovvayıoyn that discussed the views of cussion that, according to the new doxography, was Sitz im Leben of Theophrastus’ the individual doctors and thinkers on the causes of diseases. He also divides them into two groups, inside of which the sequence is mostly chronological, orientation of this work is exactly what Mansfeld, Runia and Baltussen try to reduce In the first chapter of his book Theophrastus arranges quarkol in chronological their relationship to the predecessors*, All this is unnecessary for the dialectical diswork, but it is necessary and important for the history of philosophy. Yet a historical to a minimum, if not wholly eliminate. This tendency strikingly resembles structuralism with its attempts to replace history and historical methods with structure and 15 Taig tv quota SOEatG (Alex. In Met., 72.2); Thy érévrenv dúo THY quo (Themist. In Phys., formal methods. From this perspective one could perhaps better understand massive 211.29, cf. ibid., 13.32); tag THY Aporépo quoux@v das (Simpl. In Phys., 358.12, cf, ibid., 162.8); fire ova; over-formalisation of doxography and denial of its historical character, persistent uo ÉOTL dóza (ibid., 148.28); Thy Tv kuplox qua Aryopévov dózav (id. In De caclo, 561.1); Tc claims of the methodological superiority over the outdated 19% century positivism TEL Thy Ap TOY vario dóas (Philop. In Phys., 26.23); Thy kowvhv aúvriv tv quero SEC (ibid., 89.7, cf. 94.20, 184.25, 395.18); tag THY vai Aökaz (ibid., 108.15); THY guoteov Yhoadqaoy tic Nepi Tis and firm conviction that it is a hidden dialectical structure of doxography that matters (uugdtytos S6Eag (Olymp. In Mercor. 150.28). ai Tüv quo kai tiv datpovopuriov S6Eat in Strabo most. Such an approach tends to obfuscate the fact that Theophrastus’ physical and (II, 5, 2.22-24) probably goes back to Posidonius (fr. 3c Theiler, but cf. Edelstein-Kidd 11, 135, 364). Note also Menon’s medical doxographies were parts of the historiographic project initiated by Aristotle. To the same project belong three histories of science, written by Eudemus, Abk TOV ya quiocópov and tay IDoutégyou (Euseb. Pracp. evan., XV, 340.23 Mras) and tov quatodóyev dOozat (Simpl. In Phys., 355,20). " My search with TLG in the entire corpus of Greek literature brought only one example of grow] Abe (Olymp. In Meteor., 138.29). No name is mentioned either in Aristotle or in Olympiodorus. 1 They da not go much beyond doubts expressed in the earlier literature, see: E Wehrli, Theophrast, in: H. Flashar (ed.), Die Philosophie der Antike, Bd. 3, Basel 1983, 489 1. * Baltussen, Theophr. 1, 250. The same is repeated in Theophr. 11, 245. | 5 Dicls, DG, 105. * omers?”, One more part of the project was Eudemus’ History of Theology that disnot always clear; further, the author of the papyrus seems to have made some changes tn it. Still, che first group begins with Eurvphon of Cnidos (b. after 500 BC) and ends with Acgimius of Aclis, the younger contem- # See: K, von Fritz, Democritus’ theory of vision, in: E. Underwood (ed.), Science, medicine and history, Essays written in honour of Ch. Singer, vol. 1, Oxtord 1953, 83. that were arranged according to the chronology of the mathematicians and astronporary ol Hippocrates, the second, according to Manetti (1181), runs from Philolaus to Philistion of Locri and Plato. # CH Diels, DG, 103. He wrote also on Plato and Xenocrates. Besides, he may have used the material of In $ I Herachtus is placed in che second group, but is not mentioned subsequently. Anaxagoras is the only true representative of the second group, whereas Alemacon, Clidemus, Diogenes and Democritus do not belong entirely to any of them. It seems that difficulties of clear-cut systematisation prompted Theophrastus Aristotle's monographs on the Pythagoreans, Xenophanes, Alemaeon, Melissus, etc. (D. L.. V, 25). 4 Mansfeld (n. 42), 65. CL. "I submit that in this work the tenets were the main issue, not che persons who subscribed to chem ... This, at any rate, holds for Aétius and the other sources ...” (ibid). Luckily for us, not to use the simplest, i.e. chronological, principle. Clidemus’ position between Anaxagoras and Diogenes is the only chronological indication for this obscure figure, on which he was dated by Diels. CH. Baltussen, Theophr. everything that holds for Aétius, holds also for Theophrastus. Il, 16. Mansteld’s reconstruction of the very peculiar “présentation chronologico-systématique” (Aristote 58 1. et la structure du De sensibus de Théophraste, Phronesis 41 (1996), 158--88) is not convinciny. 5% MeDiarmid (n. 3), 89; von Kienle (n. 29), 62 ff. * D. Manetti, “Aristode” and the role of doxography in the Anonymus Londinensis, in: van der Eijk (n. 10), 102, 1184. Due to the lacunae in the papyrus and the high number of the unknown names this order # G. M. Stratton, Theophrastus and the Greek physiological psychology betore Aristotle, London 1917, # I. Zhmud, Eudemus’ history of mathematics, in: W. Fortenbaugh, I. Bodnar (eds.), Eudemus of Rhodes, RUSCII vol. 11, New Brunswick 2002 (in press).

cusses the doctrines of the Heoddyot in their chronological sequence ®. In spite of the significant differences between the approaches of the individual Peripatetics to their specific material, it is history and not dialectic that unites various parts of the project into the one meaningful whole”. | That Theophrastus’ book was devoted to the quuixot as a specific group is confirmed by the fact that it deliberately ignores thinkers belonging to the other “professional” groups: teoddyou, nadnurtıkol and iatgot, Immediate reasons for such separation are clear: ideas of the Veokóyot were reviewed in Eudemus’ History of Theology, opinions of the doctors in Menon’s ‘latouxt) ouvayuryh, discoveries of the portypatiKkot in Eudemus’ History of Geometry and History of Astronomy. But whereas all these works except for the History of Theology were inclusive, i.e. they contained the names of some guoıkol, duo dOEuL seem to be exclusive. There are no VeoAöyor even in the section megt Beod (Aét. I 7), nor doctors in the embryological part. The absence of the pad qpatiKxoi in the astronomical part appears even more striking, since Hippocrates of Chios and his student Aeschylus, Meton and Euctemon, Archytas and Eudoxus did express their doxai on this subject. The omission of all these names can not be explained only by the state of our sources, it must result from deliberate choice®!. Simplicius twice refers to Theophrastus’ book ®uowî iotopia (fr. 228B, 234 FHSG), and once simply to ‘lotopia (fr. 226B). Usener and Diels regarded this book not as “physics”, but as “history of physics” and, therefore, identified it with the Puoux@v d6§at. That is plausible, for Simplicius, usually very pedantic in giving the correct titles, never refers to the Pvouxiwv dózaL. It seems that the latter book was known to him only as buouwxi iotopia®. If Dvouch intopia really meant “history of physics”, this would be a strong argument for the historical character of Theo- | 5 Tav zzgl 16 firiov ioroplug a’-5’ fram Theophrastus’ catalogue (251 N° 2 FHSG), which has to be identitied with Eudemus’ History of Theology (fr. 150 Wehrli). See: Usener (n. 40), 18. 231 phrastus’ doxographical work. But Steinmetz. denied that quoi intopiu could possibly mean “history of physics”, it must be “physical inquiry”. Accordingly, he identified this work with Theophrastus’ Physics and assigned to it most of the fragments of Diels’ Puan dózar, Theophrastus’ doxographical work was about to disappear, whereas his Physics became an alternative source of the doxographical tradition. Although Usener and Diels were basically right on this point, they did not adduce all the possible evidence. It has to be reminded that iotopiu appears in the three titles of Eudemus” histories of sciences, Feuptetoucd iotopia, ‘Aotyohoyuci) iotopia, "Apt iotopia, Simplicius quoting two of them. Eudemus’ Fempetourty ("Aotpokoyıki)) ¡utopía means not “inquiry into geometry (astronomy)”, but only “history of geometry (astronomy)”, which is clear to everyone acquainted with the content of these books, including Simplicius*. This raises the possibility that in referring to Theophrastus’ buoi] Lotogla he was aware that it constitutes a parallel, even if not exactly parallel, to Eudemus’ histories of mathematies*”. There can be no doubt that Simplicius’ references to Puouh iotogia have to be set apart from his more numerous references to Theophrastus’ Physics, which he consistently calls dvoid, Eudemus’ Physics can serve once more as a good parallel: Simplicius gives almost a hundred quotations from this book and always refers to it just as Puoi. The differences in the content between Theophrastus’ theoretical Physics and doxographical buoi] lotogía, as quoted by Simplicius, are no less evident. All three fragments from vox) iotogia concern the views of the Presocratics”, whereas only one of the ten Simplicius’ references to the Physics mentions a name (fr. 238 FHSG, Democritus), the rest is purely theoretical. All this makes impossible any attempt to consider Physics an important source for the doxographical tradition. On the other hand, it is perfectly possible that Theophrastus’ Physics contained some doxographical passages, as Aristotle’s and Eudemus’ Physics did, which means that Diels was wrong in assigning Phys. op. fr. 13 (= fr. 238 FHSG) to the voy d6Eut just because it mentions Democritus. The fact remains, however, that this is the | Y “Aristotle’s own great achievement in the field of history during his later years and the parallel works of his disciples organized by him show that the investigation of the detail occupied his mind on a large scale ... This sort ot historical interest cannot be explained any longer as an outgrowth of his dialectical method ... We must not separate Aristotle’s interest in the history of philosophy from his historical research in all these other ‘4 Steinmetz (n. 4), 3354. He was followed by Mansfeld (n. 11), 253 and (n. 23), 26, Runia (n. 39), 117, 129, 137, and Baltussen (Theophr. 11, 241). fields of civilization” (W. Jaeger, Rec., AJP 58 (1937), 354). $ Later doxography adds to the original stock opinions of the suethyatixoi and iatoof (see ‘mathematici* and ‘medici’ in Diels’ index); Geokdyou are represented by Orpheus, Homer, Hesiod, ete. Opinions of Hippocrates of Cos and Polybus also were included after Theophrastus, whereas Alemacon was considered * Er, 140, 148 Wehrli; H. Gottschalk, Rec., Gnomon 39 (1967), 20. On iotopia as “history” sec, e.g.: Arist. Rhet. 1359 b 30f., 1360 a 30-36; Poet. 1451 b 1-7, 1459 a 21-24; Anaximenes (FGrHist 72 F 3, 9). Ephorus’ universal history was entitled *lotovia; Theophrastus’ student Praxiphanes criticised Thucydides in his Hegt by him as queinéz, Cf. Diels, DG, 232; Mansfeld (n. 39), 3058-59; 1). Runia, The Placita ascribed to doctors iotopierz (Ir. 18 Wehrli). ‘Theophrastus’ praise of Herodotus’ and Thucydides’ style (fr. 697 FHSG) may come in Aétius’ doxography on physics, in: van der Fijk (n, 10), 244-45. from his Heol iotooiig, see: Regenbogen (n. 13), 1526. _“ Eudoxus’ opinion on the Nile's floods (DG, 386.11.) obviously got inte the doxographical tradition after Theophrastus (cf, DG, 228f.; Eudox. fr. 287-288 Lasserre), as mali as the reference to Eudoxus and Aratus (DG, 347.21 f.). ° Usener (n. 40), 26f.; Diels, DG, 102. Among other considerations they relicd on Priscianus, who distinguished between Theophrastus’ Naturalis historia and Naturalis anditus (no. 137 Ib, 5b FHSG). "* Sharples (n. 42) 12 notes chat there is no overlap between authors who cite Heowci iatopice by name and those who cite brauxdi dóZa by name. See Diels’ index to Simplicius, p. 1447. | % He gives a long wordy quotation from the History of Geometry (Ir. 140 Wehrli) and three of the seven quotations preserved from the History of Astronomy (fr. 146, 148-149 Wehrli). 7 jorogikei dvaypagai mentioned by Simplicius immediately alter his account of the Presocratic principles (In Phys., 28.33) imply above all Theophrastus’ work. “8 Fr, 143, 144b, 146, 149, 153a, 153c, 176, 238, 279, 298a FHSG. & Fr. 32-34, 37b, 43-44, 50, 59, 62, 75, $1, 85-88, 93a, 101, 104-105 Wehrli. 72 Anaximenes (fr. 226b FHSG), Anaximander and Anaxagoras (fr. 228b), Parmenides (fr. 234).

only fragment of Theophrastus’ Physics, reliably his Crities Philologus 145 (2001) 2 identifiable either by the title or b the content, where a name of any Presocratic appears7!, Stobacus’ quotation”, But even if Theoph rastus’ Physics had as many doxographica l by the Hua dba. Among many important differe nces between doxogra shical passages from Eudemus’ Physics and fragments of Theophrastus’ PR ad hic work, the following are the most conspicuous. Eudemus never gives 1) che. Ki nymic of the person in question, 2) nor his birthplace, 3) nor the name of his ihe contrary, in his historical work Teta. HETOUKT] Lotopía Eudemus gives both the birthpl ace and teacher’s name of almost al the mathematicians and treats them strictly in chronol s he vue N It seems, therefore, that buctki iotogi as . ogical order (fr. 133 Wehrli. . an alternative title to Ovo KxOv dela was not so erroneous, | | IV The predilection of the new doxography for a systema manifest in the repeated criticism that, in tic approach is especially his Vorsokratiker, Diels arranged material chronologically and according to the school s: This | ve of of presentation on tends te to obfuscate an i act that ... our varhier sources present the important aspect of the evidence thus cut up, doxai of che Presocraties in " Theretore, Diogenes Laertius’ careless note that viz. the clusters” Theophrastus sets out in his Physics the opinions almost all physicists (IX, 22 = 227D FHSG) is entirely wrong, Usener reter to the ais dos, the Epitome The same critique is to be found in Runia 15 years later: except for the murs: digressions as Eudemus’? (which in itself seems rather improbable), this mens hardly change anything. Neither Aristotle’s, nor Eudemus’, nor Theophrastus’ doxographical passages could have constituted the doxographical genre, as it was formed 4) nor any chronological indications. On the 233 of and Diels écrasés took this nu si To put it somewhat paradoxically, the doxographical material, though noc in itself fragmentary, became such because it was linked up to the study of individual philosophers ... This method is quite inevitable and certainly not wrong in itself. The problem is that one can very easily lose sight of the content and method of the original doxographical sources”, It is not without interest - at least for a future historian of the modern intellectual history — to note how in the course of the 20% century the term “original source” or “early source” has taken on an opposite meaning and begun to denote not Anaxagoras’ treatise, and not even Theophrastus’ compendium, but Aétius and his epitomators. The cart is obviously put before the horse: since for some later compilers (but certainly not for Theophrastus!) the individual Presocratics happened to be just “name-labels” attached to the different doxai presented in clusters, all those who arc interested in the first place in Presocratic thought distort these “original” doxographical clusters by trying to connect the ideas with their authors. But had Diels been really so impressed by the technique of the late compilers, and had he devoted his efforts to the study of the problems his critics consider to be the most important, we would hardly have benefited from his Fragmente der Vorsokratiker. The other frequent point of criticism, namely that Diels ignored Aristotle’s role in the formation of the doxography, scems to me equally untenable. In his DG, Diels was not writing a history of “influences” but was looking for a common source of the doxographical genre. He has found this in Pvotkd@v d6Ëa and convincingly connected it with the later (existing or perished) works. In other words, he was doing that very necessary Quellenforschung which now tends to be so disreputable”. Although Aristotle did not write any purely doxographical works, Theophrastus’ dependence on his teacher did not escape Diels: he finds many parallels with Aristotle’s doxographical passages and refers to Zeller’s article on this matter”, So the problem is not that Diels was unaware of something, but that he failed to demonstrate his knowledge in a form acceptable to his critics”. I cannot help thinking that had he collected all the relevant Aristotelian passages already in the DG, and not in his later of which Diogenes mentions in IX, 21. CL Galen's reference to shee Epitome (fr. 231 FISG). Doxographical fragment on the origin of the sea (fr. 23 Diels = Ir. 221 FHSG: cf. Ad II à 18) also comes from the brani dz, see: Sharples (n. 42), Mei DES e sos NE vErgage V ap s plicius 219. | (fr. 232 FHSG). One cannot expect the da crenc BER the book titles, Theophra stus’ reference to Xenaphanes Physics or from the brain adzut. See: Steinmetz (n. 4), 334 ff, R, We Sharples, heavens, in: J. Wiesner (ed.), Aristoteles Werk und Wirkung | 7 i an cf 14, 53-54, 60, 65, 67, 75, u il Si y ni among can be either from Theophrastus on the , Bd. 1, Berlin 1985 577-93; M. 147€; Runia (n. 39), Peri Rose same level of accuracy 78, 82,89, agmen: STN via A lexander, Parmenid es’ fi | ven ee ee name may well be his addition, but since - Doxographical remarks i DG ... as cause to effects, we may say that the effects have persisted while the cause has largely been diseredited, or even forgotten” (id. (n. 8), 17). See also above, n. 32, ° Lione Niki (DG, 226~29), which is far more relevant to the shaping of the doxographical tradition than Aristotelian dialectic. To be sure, at chat time Diels was not sure about its authorship, Cl, J. Partsch, Des Aristoteles Buch “Uber das Steigen des Nil”, ASGW 27 (1909), 555 n. 4. Cf. below, n. 121. quasi-biographical mode of presentation, though ” “He also failed to acknowledge the influence of Aristotle, To be sure, most Aristotelian passages dealing | with the Presocratios are to be found in the later Fragmente der Vorsokratiker, but his decisive role in shaping the tradition had fallen by the wayside” (Mansfeld (n. 8) 147). Cf. ibid., 165. based 7 à ((too) ona vo) clear hypoth. is cle sts concern concerniing the transmis sion, effective issi Ffecti ly obscures its own foundations and also inIbits access to the original sources ” “Quellenforschung, just as psycho-analysis is an heirloom of 19% century positivism” (Mansteld (n. 10), 345); “Quellenforschung is almost a dirty word today” (id. (n. 39), 3064); “che technique of Quellenforschung on which Diels’ researches are based has today fallen in bad repute ... I... we relate Quellenforschung vo the * Diels, DG, 105-6. He also compares Aétius’ chapter on the Nile floods with Aristotle’s De minidais Zvinv 6 Happuvidos yvemuyro s (fr. 372), very scarce: Gottschalk (n. 65), 20, ? Mansfeld (n. 11), 23. Cf. id. (n. 23), 25: “Diels' PT 110-11 1, 118 Wehrli. the nearly 30 Iragments Pepe non-historical works also were TM Runia(n. 10), 37. See also: Baltussen, Theophr. 1,5. themselves”,

L. Zumun, Revising Doxography: Hermann Diels and his Critics Philologus 145 (2001) 2 Vorsokratiker, he would have created even more grounds for criticising x ; “uo Ric book as “imposing » . si 5 } 2 > À his 854 page a and his Prolegomena as “labyrinthine”. Anyway, had he even started from Aristotle, he could be accused of failing to recognize the importance of the carlier writers! The following quotation shows that my last guess is not just a polemical exaggeration: Diels’ influential works provide important epistemic obstacles to the attempt to investigate whether not before Aristotle, or before Plato, doxai were already inventoried in a systematical or way, or successions were listed, Attempts 10 go beyond Diels have therefore brought only partial results, or have nat advanced far beyond the status of postulates®, I fail to see exactly which “cpistemic obstacles” Mansfeld has in mind, if Snell in 1944, and after him Classen in 1965 have found an important predecessor of Aristotle, Hippias, whose book Patzer has reconstructed in 19868!. I consider their results to be no less significant than Mansfeld’s, whose thorough study of the pre-Platoni c doxography after all relies on them, and not vice versa. And if these results are only partial, such is the nature of every scholarly explanation. Again, it is the same ideal of an all-embracing scholarly theory, which scems to be a starting point of many excurses of the new doxography into Wissenschaftsgeschichte. Mansfeld and Runia open their Aétiana with a long prehistory of the Aétius’ hypothesis (p. 1-63), starting with the early 164 century and noting every coincidenc e of Diels’ complex theory with every remark that leads, or could lead, if pressed in the same direction as Dicls. But does this really mean that he was less independen t than we used to believe? Every theory is a lucky combination of previous ideas, so in spite of the impressive number of Darwin’s predecessors, there is only one Darwin. While using so much space to highlight the most distant corners of Diels’ intellectual background, Mansfeld and Runia do not say anything about the immediate conditions under which the DG were written: that Diels was a teacher in a provincial Gymnasium who had to teach various subjects for 22 hours weekly, that he worked 18 hours per day, that many editions he had to use were unreliable while manuscripts were not easily available, and so on. The only passage where they touch one of these facts with a very promising note: “we should not be quick to point the starts 1] 23 Le I looked many times through the text and footnotes of the Prolegomena and did not find anything parsimonious and obfuscating in the hundreds references by Diels to the scholarly literature“. And is it reasonable to surmise that a reference to Budé or Jonsius could diminish Diels’ academic credentials? Of course, 1 can be as subjective as Mansfeld and Runia are; the problem is, however, that they have not yet brought any concrete example of what exactly Diels had tried to conceal and to obfuscate, so that we are not able to verify their accusations. In those cases when we are, they seem to be quite unfair, for example, their unconcealed wish to stress Diels’ dependence on his teacher Usener, Here is only one typical example. It is not correct that Diels’ edition of the Puatkdy d6Ea “slavishly follows” Usener’s editionTM. Diels retained Usener’s numbering, but in all important respects there is nothing here that even distantly resembles “slavishly following”. Diels corrects Usener’s text frequently, he does not agree with his conjectures or proposes his own, his apparatus is four times lager than Usener’s, which in itself shows how seriously he took the editor’s task. Diels marks with the special typeface the words that belong, in his view, to Theophrastus; some of Theophrastus’ fragments are longer in Diels than in Usener (fr. 2, 5, 12, 23), some are rearranged (fr. 6a); there are fragments not represented in Usener (fr. 5a; fr. 19 was omitted in Usener due to the typographical error). Although Diels respected his teacher and trusted him, he was hardly “blinded by loyality to Usener’s hypothesis”, as Runia says elsewhere.8 Rather, he did not want to demonstrate independence at any price which, as it seems, is exactly what Runia does. When looking for the criteria for separation between material of Arius Didymus and Aétius in Stobaeus’ compilation, he stresses: It is thus of vital importance to determine various criteria that allow the two chief sources to be separated. In our view the most significant of these are eight in number. Obviously our list takes as its point of departure the list of ten criteria which Diels outlined in his rigorous and admirably succinct analysis. It would seem better, however, to draw up a new and revised list, in which his results are incorporated rather than repeat his list more scholastico ... Unlike Diels we shall exclude any considerations that are based wholly on content (i.e. Diels’ last three criteria). accusing finger here”, and here is the continuation: * Cf. Baltussen's economical references to his predecessors, In Theophr. I, 13-14 n. 5 he says that the purpose and function of the DS as a whole have remained unstudied ever since Diels printed (!) it in the DG, Stuck in a school-teaching job which he loathed, Diels was determined co establish his academie credentials as impressively as he possibly could. Hence the erudite Latin of the Prolegomena, its complex and circuitous structure, the scanty and not seldom scathing references to previous and contemporary scholars, For this reason too, we may surmise, his very considerable debts to his predecessors ... are rather successfully concealed *, and mentions two monographs on this subject (J. Beare 1906, repr. 1993; G. M, Stratton, 1917, repr. 1964) only in one critical footnote. In Theaphe. I the stakes are raised: “{This} investigation ... represents the first monograph on chis text since L. Philippson’s //ylé Anthröpine (1831)”; “until recently the DS as a whole remained virtually unstudied ... Atany rate, Stratton’s commentary is insufficient” (1-2). Ch. a very different opinion: “the work by Stratton ... scents by far the best so tar as the interpretation of Theophrastus’ work is concerned” (von Fritz {n. 49), 84 n. 6). = Runia (n. 39), 115. K a * Runia (n. 39), 116. “Diels ise sich zeit seines Lebens dessen bewußt geblieben, daß Hermann Usener es Mansfeld (n. 11), 23. Cf. id. (n. 8), 25. war, der ihm die erste große Aufgabe stellte ... Y A. Patzer, Der Sophist Hippias als Philosophichistoriker, Freiburg 1986. ui Aétiana, 73, cl. Mansteld (n. 8), 168. For a similar kind of Wissenschaftsgeschichte Hermann Diels: What sort of fellow was he?, in: Calder III (n. 8), 1. und nie hat Diels aufgehört, mit ciefster Dankbarkeit von diesem begnadeten Lehrer ... zu sprechen” (O. Regenbogen, Hermann Diels (1922), Kleine Schriften, Münsec: W. Calder IN, chen 1961, 545). + D. Runia, Additional fragments of Arius Didymus on physics, in: Algra (n. 18), 367 and n. 1.

Arithmetically, this means that Runia subtracts three criteria from Diels’ list of ten, then adds to the rest one criterion of his own and calls this “a new and revised list”, in which Diels’ results are incorporated! It is worth citing here Runia’s carlier note: “The results of Diels’ analysis were incorporated by Wachsmuth in the very structure of his edition, which therefore cannot be regarded as an independent piece of work”#7, To what extent are then the studies of the new doxography school independent in this specific sense, e.g. the first volume of Aétiana? “Diels did not easily change his mind or develop his thought”®*, It is true that Wilamowitz, to whose Erinnerungen Mansfeld and Runia refer in this case, did it much more casily, which only underlines profound differences between the two life-long friends. Anyone who has read Diels would probably agree that he always tried to build his views on the most solid factual basis that was possible in a given case, so he did not need to change his mind as quickly as Wilamowitz. It does not follow from this that he was unwilling or unable to correct his doxographical theory®. At least, Philologus 145 (2001) 2 neàt 237 Theophrast ... schrieb die erste „Geschichte der <Narur>philosophic”; (das Buch Theophrasts Über die Meinungen (doxai, daher Doxographic) der Naterphilosophen ist verlorengegangen, mit Ausnahme der Kapitel „Über die sinnlichen Wahrnehmungen*”, ee Except for one sentence, “jedenfalls hat keiner es bisher gewagt, Diels’ Arbeit neu zu machen”, nothing here suggests that the author is going to undertake a radical revision of Diels’ and his own views. In 1987, Mansfeld was still using Diels’ variant Hua dE, yet by 1989 he had changed his mind,” and in 1999 Runia points the accusing finger at Wehrli (1983!) by marking this variant of the title with the word sic, as if it were a grave or curious mistake.” In 1987, Mansfeld was still pleased at finding at least one mistake in Diels: It can be shown that the doxographies concerned with Xenophanes ... provide the unique opportunity to prove Diels wrong in at least one important case, Le. not absolutely right insofar as his reconstruction as a whole is at issue”. Diels’ letter to his publisher De Gruyter written on the eve of the first World War can hardly serve as decisive evidence for such a conclusionTM. In the preface to the fourth edition of the Vorsokratiker (1922) Diels complains that, because of the higher printing costs, he cannot rearrange and improve this work as he wished and adds: Das Schicksal der Doxographi, die vergriffen sind, aber die Neubearbeitung in kürzerer Form, die dringend nötig gewesen wire, nicht haben finden können, verpflichtet mich gegenüber dem Verlage dieses vorliegenden Werkes ... zu dem herzlichen Danke. If I correctly understand this, Diels in fact wanted to have a new edition of the DG, although he hardly felt the need for its complete revision. After all, this is not such a bad thing, for had he changed his mind as easily as his critics do, we would have at least three substantially different versions of the development of the doxographical tradition. Indeed, it would be unfair to say, and even more unfair to repeat on every occasion, that Mansfeld’s Vorsokratiker slavishly follows Diels and that, like Diels’ work, it is person-oriented and not doxai-oriented, which eo ipso distorts the original context of the doxat. The character of this book fully corresponds to its task, so that such reproaches would be beside the point. But it seems worth recalling that in the preface to this It turns out, however, that the main thesis of his paper, namely that the doxographical tradition offers in this case not Theophrastus’ report on Xenophanes, but Theophrastus’ own view, is wrong. This was shown in 1992 by no one else other than Runia, who concludes that Diels was right after all in this issue”, In 1986 Mansfeld notes: Steinmetz’ suggestions “that the fragments in Simplicius attributed by Usener and Diels to the doxographical work are from Theophrastus’ Physics no longer have the same appeal for me”, but three ycars later he fully accepts them”. In 1992 Runia insists that it is “methodologically misguided” to try to connect various doxographical sources by means of a stemma, as it is implied in Diels’ naive Einguellentheorie”, but in 1997 such a stemma appears in Aétiana (p. 328). The problem, however, is that in the DG there is no stemma for the entire doxographical tradition, neither for its earlier period, nor for the later. The only stemma to be found here concerns the textual transmission of Ps.-Plutarch's Epitome, including its manuscripts and the sources that depend on it”. In contrast to Runia’s 1992 reconstruction and to the diagram in Aétiana (p. 81), Aétius, Vetusta placita and Theophrastus do not appear here. book we find exactly those views that Mansfeld and his school have been attacking during the last 15 years: 1, Mansfeld, Die Vorsokratiker, Stuttgart 1983, 30. 2 CT. Mansfeld (n. 11), 147 and 262 n. 49, ” Runia(n. 10), 46 n. 54. 4 Mansfeld (n. 11), 149-50. 1? Runia (1. 86), 366. Aétiana, 106, repeated in: Mansfeld (n. 8), 163. * Ege he changed his mind on the dependence of Achilles Tatius on Aétius (before it was suggested by G. Pasquali, Doxographica aus Basiliusscholien, NGWG 1910, 221 n. 1), which is one of the few important points of divergence between Diels' reconstruction of Aétius’ work and that of Aétiana. “© Quoted in: Mansfeld (n. 8), 162, Cf. Runia (n. 39), 117 n. 14. “ Runia (n. 39), 129. “+ CF, Mansield (n. 109,72 n, 16 and 238 ff. 7 Runia (n. 39), 137. On p. 136 Runia gives a diagram, which is “entirely in the spirit of Diels”, "* DG, 40. Liceat mihi hac loco, quoniam varios Plutarchi libelli casus et memorias ab ipsis incunabulis per actatum gentiumque seriem ad recentiora tempora decucurrt, stemmatis artificio longiorem disputationem breviter conprebendere et illustrare (ibid., 39).

Reviewing T. Göransson’s book on Middle Platonism, Runia approvingly notes: 239 thing historical in Theophrastus’ doxography is methodically suppressed, and every- “Seldom has a scholar shown such a sharp eye for the weakness of other scholars’ thing systematic and critical is forcefully connected with dialectic, as if it were the only arguments”. | am quite sure that a scholar with a much sharper eye than mine could possible source of systematisation and criticism. Sometimes one gets the impression easily double or triple my selective list of passages illustrating how eagerly and that Theophrastus was writing his work with the Topics in his hand and that otherwise he would be unable to write it at all. For instance, in the DS he frequently criticises one thesis from various points of view. Now, where else did he learn this remarkable and sophisticated technique of critique if not in the Topics? It is fairly natural therefore to argue that by applying this technique he “is following the advice of Aristotle expressed in the Top. © 8 to formulate many arguments against one point” !%. The notion of dialectic as used by the new doxography school and its supposed role in structuring the ancient doxography scem to me the weakest point in the whole debate. To be sure, the expansion of dialectic from its original home in the Topics to various areas of Aristotle’ philosophy and the tendency to make dialectic his most important philosophical method are dated at least fifty years earlier. But joining this trend of research, this school expands dialectic much further, both thematically and chronologically, from Aristotle’s philosophy and the doxographical passages in his theoretical treatises! via Theophrastus’ compendium to the late doxographers. At the same time, it reduces the notion of dialectic to a few easily recognizable signs like diaeresis, while ignoring the other, more important and characteristic features of dialectic which are not to be found in doxography. Any detailed analysis of the “dialectical doxography” would go beyond the limits of this paper, so I have to be more apodictic here. There is a very revealing parallel in the recent scholarship: almost simultaneously with the expansion of dialectic occurred the systematic application of the Posterior Analytics to Aristotle’s zoological works. quickly Diels’ eritics change their views. Yet I would like to draw attention to one more fact. Critical remarks on naïvité of the Zingiellentheorie are not to be found in the Aétiana or in the later publications on the doxography, which plainly avoid mentioning this word. Docs this mean that working closely on Aétius’ tradition, which represents a brilliant example of the successful Einguellentheorie, Mansfeld and Runia feel uncomfortable repeating their previous criticism of this would-be antiquated method '®? Whatever the reasons for this change are, one is tempted to note beth the differences with Diels and the similarity of the new doxography to the ancient Greek one: The doxographical method was highly flexible and able to be adapted to a large number of different situations and usages 101, Vv The complex structure of the ®votk@v dÓEuL guarantees the unity of the doxographical genre and allows to recognise its dependence on the original source. Now, we come to a very important point: what was the goal and structure of this book? The traditional view is that it was a Peripatetic history of philosophy, systematic and critical !®, which should not be confused with the modern history of philosophy but can be compared with it, being in fact its starting point. Quod non, objects Mansfeld, — this compendium was written for dialectical purposes and, respectively, has a typically dialectical structure: Aristotle converted Plato's urbane approaches (and Hippias’ and Gorgias’ presentations) into a dis- AedLmo Intermediate results of this last trend were summed up and critically assessed by G. Lloyd: “results so far obtained are disappointing”; “interpretative anarchy”; positions are changing in a couple of years; attention is constantly drawn to the need for future work1%, Quite similar features are discernible in the application of the cipline, viz. dialectic, which follows a set of specific rules, and it is from this discipline that doxography Topics to the buouk@v dóEa: the same disappointing results, interpretative anarchy. qua genre stems!0, constantly changing views and references to the need of further work. The reason It sounds paradoxical, but struggling against Diels’ idea of the Puarxiv dóEaL as Urquelle of the later tradition, the new doxography substitutes its own Einquellentheorie with Aristotle’s Topics as fons et origo of the entire Greek doxography. Every- 14 Baltussen, Theophr. II, 224. 1% On previous dialectical interpretations of Aristotle’s doxography sec: E. Berti, Sul carattere “dialettico” della storiografia filosofica di Aristotele, in: G, Cambiano (ed.), Storiografia e dossografia nella filosofia antica, Torino 1986, 101-125, Berti believes that it was Cherniss who showed the dialectical character of Aristotle’s historiography of philosophy (ibid., 105). Cf. Jacger (n. 59), 350. ” D. Runia, Rec., Vigilia Christiana 51 (1997), 107. 109 Baltussen’s (n. 14), 237 continuous attacks on Eingwellentheorie rest on an erroneous understanding of what it implies. Although Diels regarded duo d68et as the ultimate source of the doxographical tradition, it was never his idea that every lemma in Actius directly goes back to Theophrastus. 10% Runia (n. 10), 39. le G. E, R. Lloyd, Aristotle’s zoology and his metaphysics: The status quaestionis. A critical review of some recent theories, in: Methods and problems in Greek science, Cambridge 1991, 373-396, 177 Sec, eg: Diels’ “hypothesis is currently being revised and in need of still further revision, so the account which follows, though still preliminary in the sense that chis revision is not yet completed, will in part go beyond Diels” (Mausteld (n. 23), 26). Baltussen (Uheophr. II, 243) is sure that Usener-Dicls’ hypothesis is 102 “Der Aufbau scheint nach Sach- und Problemkategorien geordnet gewesen zu sein, innerhalb deren bad, but “what should replace the hypothesis is not casily stated and will require further investigation”. Cf. his sowohl «lie zeitliche Folge als die angeblichen Schulzusammenhänge bestimmend waren” (Regenbogen explanation of why he excluded a chapter on the relationship of the DS with Aétius: “This is a measure (n. 13), 1536). inspired both by the need for further thought on the relevant sources and by the recent developments in the 103 Mansfeld (n. 31), 24. study of doxography {Aéuus)” (Theophr. 11 234).

for this similarity is obvious: systematic application of the logical procedures in doxography is a priori as inconceivable as in zoology. As for the accidental application, it is categories doing and being affected”!!*, The question, however, remains: did the 240 241 Presocratics develop various theories on the solar eclipse or not? And is not it only natural for a person, who was interested in giving an overview of such theories to worth quoting Aristotle himself: place them under the heading regi Exdetpews fhiou? All these are questions to which Hence evervbody, including laymen, makes use in some way of dialcetic and peirastic; for all to some I did not find any clear answer in the numerous publications on “dialectical doxoextent undertake ro test those who profess to know things Soph. Ref. 172 à 30). graphy”. Is diaeresis really so prominent in dialectic as to be its main representative in doxography? And was not Aristotle's later attitude to diaeresis rather critical 102 Theophrastus surely tried to make similarities and differences between the Presocratic doxai as clear as possible, but this is quite understandable in the context of the history of philosophy, without any appeal to dialectic. In its loose form diaeresis is a method of classification, perfectly applicable to different material. If there is something in doxography that looks like diaeresis, or would look so if we restore some missing links, it does not mean necessarily that doxography qua genre stems from dialectic, because the very equation ‘diaeresis = dialectic’ is more than doubtful. What is more, such restorations often appear to be quite artificial 1%, sometimes they do not square with the material at all!'%, Where diaeresis is in fact present, c.g., at the beginning of the DS (above, p. 228), it is both evident and in a sense trivial as any similar division (realists and nominalists, materialists and idealists, etc.), but the postulate of a complex and hidden diacretical structure of the entire doxography is unsupported by Tun, D$aI “are specific applications of the general topic tepl kivijoemg in the sense of the 30, Rhet. 1359 b 10ff.). Respectively, those contemporary experts in dialectic, who study the Topics for its own sake and not for the sake of doxography, are very restrictive concerning its possible applications in philosophy and science!!5, Primavesi, for example, maintains that the Topics describes a method of a dialectical syllogism, i.e. of a compulsive argumentation in yutvacia, that dialectic presupposes real, and not just imaginary, dialogue between two persons, with strict rules of questioning and answering, etc.''®, Obviously, we would look in vain for that kind of thing in the Dvoikdy dOEUL, and that is why the new doxography never attempted to find it. Among those things that it has found, the following are worth mentioning. First, Aristotle remarks casually that when preparing for dialectical discussions we have to make selections of opinions from the existing literature, including the doxai of individuals, e.g. Empedocles’ opinion that the elements of bodies are four (105 a 34f.). Now, if something is used or just mentioned in the Topics, it does not necessarily become our evidence !", What is the difference between EvöoEa. as the basis of dialectical syllogism, and dózca, as the subject of doxography? Presocratic d6Eat include a lot of Evöoku, but also many things which Aristotle would surely have qualified as údoZa !12, Mansfeld regularly refers to the Categories and Posterior Analytics in his explanations of the dialectical nature of doxography "3. Does it mean that both Aristotelian logical categories and his theory of the scientific syllogism are part of his dialectic? It is possible, of course, to interpret some chapters of the Pbvaikov d6Eu in terms of Aristotle's categories, suggesting that the chapters met tvombv nAlov and meg Erkeipewg HALov Aristotle certainly had a more limited conception of what dialectic is, for in his works we find a sharp distinction between dialectic and philosophy (e.g.. Top. 105 b dialectical. Aristotle’s advice reflects his gencral attitude to the predecessors and his readiness to take their views as the starting point of any research in which he was interested.'!7 Second, there is the well-known passage that dialectic can be of use for An philosophy as a way to doyai of all the sciences (101 a 34f.). Disregarding the fact that this passage remains isolated, for in the Analytics Aristotle found another and better way to Gpyal 118, one can ask, whether doxography has anything in common with che search for dpyat. This question was raised by Baltussen and here is his conclusion: [Theophrastus’] evaluation of propositional archai ... can only be called a dialectical procedure in a restricted sense. Dialectic proper was defined as the reasoning from provisional archai to the foundations of a certain discipline. Such a form of reasoning does not materialize in DS. 1 H.Cherniss, Aristotle’s criticism of Plato and the Academy, New York 1962, 27 ff. 107 See, ey: |. Mansfeld, Chrysippus and the Placit.e, Phronesis 34 (1989), 314 ff. © Mansfeld's idea that Simplicius’ application of diuiprow to the chapter mpi Goze of the Brandy 14 Mansfeld (n. 42), 93. 1 The best recent introduction is: O. Primavesi, Div Aristotelische Topik, München 1996. See also: dOze goes back to Theaphrastus (Mansfeld (n. 11), 243 ff.) contradicts both Simplicius’ own words (In Phys., D. Devercux, Comments on Robert Bolton’s “The epistemological basis of Aristotelian dialectic’, in: D. De- 22.20-21) and the actual arrangement of material in this chapter. Cf. Diels, DG, 104; Steinmetz (n. 4), 3381f.; vereux, P. Pellegrin (eds.), Biologie, logique et métaphysique chez Aristote, Paris 1990, 263-86; P. Slom- J. Wiesner, ‘I'heophrast und der Beginn des Archereferats von Simplikios’ Physikkommentar, Hermes 117 kowski, Aristotle's Topics, Leiden 1997. (1989), 288-303, and above, n. 56. ti For a balanced critique of this idea see: A. Laks, Du témoignage comme fragment, in: G. W. Most (ed.), Collecting fragments/Fragmente sammeln, Gottingen 1997, 257-63. 112 Inthe Topics 160 b 17-22 the views of Parmenides (nothing is in motion) and of Heraclitus (everything is in motion) are characterized as (doa, which have to be avoided in a dialectical discussion. 10 Mansfeld (n. 42), 69 ff. {6 Primavesi (n. 115), 17 ff. See also: E. Kapp, Syllogistik, RE [V A (1932), 1056; R. Robinson, Plato's carlicr dialectic, 2"! ed., Oxford 1953, 91; Slomkowski (n. 115), 911. 7 Cf. his revealing remark in the middle of the doxographical passage: “It is what we are all inclined to do, to direct our inquiry not to the matter itself, but to the views of our opponents” (De caclo 294b 7-9, Rev.Ox.Tr.). Devereux (n. 115), 284 f.; Primavesi (n. 115), 521., 57.

He still insists that “DS is dialectical sensu Aristotelico to a considerable degree”, for it meets at least two of four minimum requirements for a dialectical work, formulated by himself, viz. 1) it makes use of endoxa and 2) it exhibits dialectical arguments!', One has to object. first, that doxography makes use not of endoxa, but of doxai, and it Runia and Baltussen in many important points! Yet this necessary work can be done without excessive struggle against or for any dominance (as opposed to an ambitious but fruitless deconstruction of presuppositions, methods and results“ of Diels’ work !25). This was recently demonstrated in the exemplary article of W. Burkert on Diels’ Vorsokratiker!"*. Instead of dominance“, he speaks of „Diels’sche Zeitalter“, in which we are still living, and does not conceal his doubts about the 242 would be strange if it did not; second, that exactly because dialectic is devoid of its own subject-matter its arguments are casily applicable to any subject'. But this 243 potential applicability does not prove the actual dialectical origin of doxography. How far the origins of doxography are from dialectic shows one of the earliest doxographical overviews, so far neglected by Mansfeld. Interested in the causes of the Nile's floods, Herodotus prefaces his own explanation of this phenomenon (Il, 24-25) with three opinions of his predecessors (Il, 20-23), viz. of Thales, Euthymenes of Massalia, and Anaxagoras (it is hard to say, whether their chronological order is accidental). Although he does not call them by names, they are casily identifiable possible replacement of Diels’ Vorsokratiker with a new, methodologically refined from the later doxography on this topic. Herodotus’ overview reveals some patterns characteristic for the Peripatetic doxography: 1) physical problem, 2) short references to the relevant opinions of the predecessors (two of them are quatkot) followed by 3) criticism of the first and the third doxa from various points of view (the second is bluntly dismissed). It is only natural that all the three doxai plus that of Herodotus himself appear in the doxographical part of Aristotle’s De inundatione Nili (who, in his turn, brings against them physical, and not dialectical arguments) and later in Aétius (IV, 1)!2!, As for the origins of doxography, it seems more than symbolical that this doxographical overview comes from a historical work written at doxography). It is far from certain, whether all these texts or just any of them will least 80 years before the Topics. edition. By showing that Diels’ reconstruction of Aétius was right „almost beyond expectation“!25, Mansfeld and Runia proved once more that this part of the scholarly tradition rests on a solid foundation. To be sure, „Diels’ Arbeit neu zu machen“ hardly means only to show where he was right or wrong. Taken seriously this would imply new editions of the doxographical texts edited by him both in the DG (among them Aétius, Arius Didymus, Theophrastus’ De sensibus) and later (Menon’s medical appear in the following volumes of Aétiana and will forever replace Diels’ work. The Wellcome Trust Centre for the History of Medicine London NWI IAD Great Britain Abstract This paper examines some important points of Diels’ doxographical theory that have recently come under attack by his critics, J. Mansfeld, D. Runia, and H. BalVI tussen. It can be shown that Diels’ theory, without being impeccable, can easily The value of a scholarly work can be estimated from the extent that it makes unnecessary all previous, and stimulates all subsequent, studies in this field. From this point of view, Diels’ Doxographi Graect deserves as highest praise as his Vorsokratiker, even though the former attracted so far much less attention than the latter. Instead of repeatedly asking “Why did Diels not close the field of doxography?”, we have to be grateful to him that he opened it for us. This does not mean of course that there is only one god of doxography - Theophrastus, with Diels as his prophet, in whose scripture there is nothing to change or to add. I do not have any doubt that Diels’ reconstruction of the doxographical tradition, both the earlier and the later, can be corrected, improved and developed further, as has already been done by Mansteld, withstand this criticism, whereas various attempts aiming at radical revision of this theory are mutually contradictory and not supported by the evidence. 1? Diels was very far from considering his reconstruction infallible: “Daß in dem komplizierten Geflecht der statuierten Abhängigkeitsverhältnisse und bei der notwendigen Scheidungen von Überlieferungsgruppen manche Linie zu einfach durchgezogen, manche Trennung zu scharf durchgeführt werden mußte, hat Diels selbst am besten gewußt: Nur so war es möglich, erst einmal eine Durchleuchtung und Ordnung des riesenhaften Materials zu erreichen” (Regenbogen (n, 85), 545-46). 2 Baleussen (n. 14), 227. 14 W. Burkert, Diels’ Vorsokratiker. Rückschau und Ausblick, in: Calder [I (n. 8), 169-97. 123 „Die Actios-Rekonstruktion der DG ist von Mansfeld-Runia (Aétiana) fast über die Erwartung hinaus bestätigt worden“ (Burkert, in: Calder INT (n. 8), 165). Cf. a generally sympathetic review of Aötiana: „Mansti Baltussen, Theophr. 11, 229, 232. For the kind of arguments he calls “dialectical” see above, p. 223. 129 “The proper subjects of dialectical and rhetorical deductions are the things with which we say the commonplaces are concerned, that is to say those that apply equally to questions of right conduct, natural feld and Runia rightly are critical of Diels’ procedure and arguments, but in the end come to the conclusion science, politics, and many other things” (Arist. Rhet. 1358 a 10-14; Rev.Ox.Tr.). crucial importance if doubts concerning this central part of Diels’ hypothesis can he laid to rest“ (M. Frede, 121 J. Balty-Fontaine, Pour une édition nouvelle du “Liber Aristotelis de Inundatione Nili”, Chronique that Diels, at least as far as this part of his thesis is concerned, was basically right ... If one chinks that this result ... is comparatively meagre given the enormous efforts made, one should keep in mind chat it is of Rec., Phronesis, 44.2 (1999), 137). lt has to be stressed, however, that before Mansfeld and Runia it was only d'Égypte 34 (1959), 95-102; D. Bonneau, Liber Aristotclis De inundatione Niki, Études de Papyrologie 9 Lebedev’s short, and mostly unnoticed, paper that attacked Aëtius” hypothesis. Significantly, only 8 of 371 (1971), 1-33; M. Bollack, La raison de Luerèce. Paris 1979, 539 ff, pages of Aétrana are devoted to the refutation of Lebedev's thesis.