Version classiqueVersion mobile
OpenEdition Books

Qu'est-ce que la philosophie présocratique ?

André Laks
Claire Louguet

B. Editer les Présocratiques

Problems raised by an edition of Democritus with comparisons with other Presocratics1

Walter Leszl

Texte intégral

  • 1 I am grateful for observations made to one of the drafts of this paper to C. Huffmann, L. Gemelli M (...)
  • 2 In some cases, as here and in the title, I talk simply of Democritus, since I do not think that Leu (...)

1I permit some remarks on how I was induced to prepare a new edition (with Italian translation) of testimonies and fragments concerning Leucippus and Democritus. This was not my intention from the beginning, for I already had some suspicion that this is a daunting task, but I intended to Write a monograph on Democritus which offered a general survey of his thought2. At that time the only available monographs of this type were those by Bailey and by Alfieri, neither of which was a recent publication. In the meantime of course the situation has changed, for, while I was slowly carrying on my long-term research, two important inclusive books have appeared in French, those by P.-M. Morel and by J. Salem, as well as various other publications. An author of a monograph hopes of course to go beyond offering an up-to-date survey, by providing some new insights on the thought of the philosopher he examines. Whether or not the monographical part of my opus will satisfy these requirements, it remains true that it is the part that was started first and was initially meant to be the only part written.

  • 3 Meanwhile I had news of an edition of Democritus to be prepared by Laura Gemelli Marciano under the (...)
  • 4 The provisory edition I made available to the Colloque in Lille is the result of such an act of wil (...)

2With the passing of time, however, I grew more and more dissatisfied with the available editions of Democritus (and of Leucippus) and got the feeling that no sufficiently reliable survey of their thought could be given without being based on a new edition of testimonies and fragments. Thus, since nobody appeared to be willing to do the job, I overcame my reluctance and decided to do it myself3. The damning side of this kind of job, which I did not sufficiently foresee, is that any time you have a look at what you have done you detect defects and lacunae, and there seems to be no end to this, except by an act of will by which one decides that one is not going to do any further work of revision4.

3In what follows I do not intend to discuss certain specific issues, such as the relationship between Leucippus and Democritus or the authenticity of the ethical fragments and the reliability of a number of testimonies concerning the latter (especially those which make a magician of him). Further issues that I leave out, though not all so specific, concern the selection of the passages, the possibility of ordering them so as to make obvious their interrelationship, and the extension of the quotations. I equally leave out technical questions about the use I make of existent critical editions in the presentation of the original texts in my collection. I try instead to concentrate on issues that can be of interest in editing a Presocratic author whose works are lost to us, considering the nature of the documentation available to us in the light of the tradition or traditions to which it belongs. There are some issues, especially those dealt with in part 3 of this paper, which would deserve a fuller and indipendent treatment, but, in the impossibility of carrying out the necessary research soon, I prefer to submit provisional results in the present form, in order to foster an increased awareness of their complexity.

I. Editing a Presocratic

  • 5 A few further observations below (these do not concern his commentary). For a general criticism of (...)
  • 6 Not all of the changes that Kranz introduces in the texts, in their translation and in their appara (...)
  • 7 Cf. the edition, with the abbreviated title Aetii plac., given by Diels 1879, pp. 273-444.

4The editions to which I made an allusion are of course the Vorsokratiker by Diels and Kranz and the Democritea by Luria. I will give some attention to the former, and less attention to the latter, for, though it is more complete, it cannot be said it has supplanted the former even among scholars who work in this field. The main reason for this is, I think, that the edition by Luria is conceived in such a way (with its rather peculiar ordering of the passages and with its repetition of the same-or nearly the same-passages, sometimes reduced to little bits, under different headings) as to render the access to Democritus rather difficult. Furthermore, wherever the passages coincide with those of the edition by Diels and Kranz, there is little improvement in the way they are presented, since in most cases they are of the same extension and are not accompanied by information on the context5. Diels’ edition (Kranz does not introduce changes from this point of view6) reflects a rather clear conception of how the texts should be presented. It is admitted that we can draw a sufficiently neat distinction between the contribution of Leucippus and that of Democritus and that, whenever the two are mentioned together in a testimony, this should go under Leucippus, not under Democritus, since the former has a (not purely chronological) priority on the latter. It is equally admitted that we can draw a sufficiently neat distinction between fragments and testimonies, so that they can be kept separate from each other and the fragments can usually be presented independently of the testimonies. Some of the fragments are associated with the titles of the writings (of Democritus) to which they originally belonged, when there are testimonies which state this connection or when it can be conjectured. In the other cases they are ordered neutrally according to the alphabetical order of the names of the ancient authors who are the source (thus passages from Plutarch come before passages from Stobaeus) or according to the internal order of the work or Works of the same ancient author from which a given passage is extracted (e.g. according to the position the work occupies in the collection of Moralia by Plutarch or the position the passage has received in a collection like that by Stobeaus). The doctrinal testimonies follow the biographical ones (including those concerning writings, style, etc.) and they are ordered (apart from the ethical ones put at the end) according to the order of presentation of the books of Theophrastus’ work supposed to be entitled Φυσικῶν δόξαι. (This is declared by Diels in the introduction to the first edition, where he suggests that the order in fact will be the following: principles, God, cosmos, meteora, psychology, physiology.) Since this order is not known in a direct and detailed way, however, the order which is actually followed is, to a large extent, the one suggested by the order of presentation of the books and chapters of the collection of Placita by Aëtius as reconstructed on the basis of the order to be found in Ps.-Plutarch’s Epitomé7. That Diels, in collecting the testimonies, largely depends on the work done in his Doxographi graeci to reconstruct the tradition going back to Theophrastus, is evident not only from the adoption of that order of presentation but also in various other ways, e.g. in the priority he gives to the passages which are supposed to be fragments from that work of Theophrastus, including the passages that come from the De sensibus – admitted to be the last book of that work –, or in printing certain passages simply and directly under Aëtius’ name.

  • 8 On these points see Burkert 1999, esp. 178-80.

5Concerning the relationship between Leucippus and Democritus, the very fact that sometimes they are quoted together tends to exclude the possibility of distinguishing their contributions, and in most testimonies concerning both a continuity between their positions is either stated or assumed, so that their separation appears to be artificial, but this (as already stated) is not a point I intend to discuss in any detail. Talking, then, of Democritus alone, what is clear is that Diels applies to him a procedure which, in so far as possible, is applied to all of the Presocratics. The obvious objective that is set, in collecting the fragments independently of the testimonies, is to reconstruct the original work or works of the Presocratic philosopher concerned. This usually entails at least to leave lacunae of some extension among the fragments, without attempting to fill them with the help of the testimonies, since these do not give the very words of the Presocratic philosopher. If too little is preserved to even get some approximation to the original work, at least the very words of the philosopher are given, keeping them in separation from any successive interpretative intervention. This favours the attempt to make the Presocratics speak with their own words (avoiding successive terminology such as the Aristotelian ὁμοιομερία in the case of Anaxagoras) and thus to understand them in their own terms. And it makes us aware, in some cases, of how little of the original has been preserved by the tradition (only one sentence in the case of Anaximander, practically nothing in the case of Thales, at best a few scattered words in the case of Anaximenes). Yet this procedure is not, as we shall see, without disadvantages8.

6As to the testimonies taken by themselves, there is a partially parallel objective that is pursued, namely, the reconstruction of Theophrastus account of the thought of each philosopher on the basis of his collection of ‘Opinions of the physicists’ (I shall return to this title later), which is supposed to be our main ultimate source of a large part of the testimonies. Since however only a modest part of this material can be taken as fragments of Theophrastus’ work (23 passages in Diels’ own edition of the Physicorum opiniones, in Doxographi graeci, pp. 473-495), while the rest is constituted by passages belonging to shortened and otherwise modified versions of this work, the objective can be pursued only by approximation. There is also the limitation that lacunae are to be found here too, since those versions are shorter not simply by summarising the original but also by omitting passages concerning this or that Presocratic philosopher.

7Manifestly it is assumed that the two objectives are parallel but not overlapping, so that they can be pursued in an independent way. The obvious reason for this is that Theophrastus’ work and the collections of placita or Epitomai which are based on it do not usually provide extended quotations from the original Works but only a synthetic exposition of the thought of the Presocratic philosophers concerned (this way of proceding is well illustrated by Theophrastus’ De sensibus). A complication however is given by the fact that not all ancient expositions of the thought of a Presocratic philosopher are of this type, since some of them offer a mixture of testimony and quotation. For instance in Hippolytus’ Refutatio his main presentation of Heraclitus’ thought (in IX, 9, 1-10, 9) is accompanied by the quotation of various Heraclitean passages. Even in the case of a testimony that can be taken as going back to Theophrastus, the main one by Simplicius concerning Anaximander, it contains the longest quotation we have of his words (the aforementioned one sentence). Hence the separation between fragments and testimonies is not always so neat (especially as it is not always clear where exactly the quotation starts and ends), and in fact it happens with some frequency that Diels, under the heading of fragments, does not restrict the quotation to the ipsissima verba of the Presocratic philosopher but extends it to the context to which it belongs. (Examples could be given not only in the case of Heraclitus but in the case of Parmenides etc., and of course of Democritus) But, if this were the only difficulty that arises in trying to provide fragments and testimonies as two independent blocks, one could maintain that the disadvantage which is involved by admitting these exceptions to the rule that fragments and testimonies be kept distinct is compensated by the advantage constituted by the neatness of the distinction itself. There are however other difficulties, which become evident in raising some general questions.

8One question that has to be raised is whether a uniform presentation of the Presocratic philosophers, by adopting Diels’ criteria (starting from that of offering fragments and testimonies as two separated blocks), is in fact the most satisfactory one. It Works on certain assumptions which may not be applicable (or equally applicable) to all of the Presocratics. One of these assumptions is that the testimonies, or at least most of them, belong to one tradition in that the sources of testimonies in fact depend (rather exclusively) on one work, identified with Theophrastus’ collection of ‘Opinions of the physicists’. Whether, and how far, this is applicable to this or that Presocratic thinker is an open question. This question, we shall see, is particularly relevant in the case of Democritus. Even if this preference for a Einquellentheorie were fully justified, it would remain an open question whether an approximate reproduction of Theophrastus’ account of the thought of the given philosopher is possible and, if so, is very useful for the purposes of a work like Diels’ Vorsokratiker, when the Physicorum opiniones presumably did not contain a full exposition of the thought of the Presocratics taken one by one, because each of its books must have been dedicated to a different topic.

  • 9 That this is Aristotles interpretation of their position is very clear in De caelo III, 1, 298b21 f (...)

9Leaving these questions aside for the moment, we have to consider another assumption that is made in admitting that Theophrastus’ work should be taken as being, directly or indirectly, the basis of the documentation that is different from quotations from the original works, namely the assumption of the adequacy of such a work for all of the Presocratics. What happens, from this point of view, to those testimonies which in fact are likely to depend on Theophrastus or in general reflect a Peripatetic approach? It would seem that, if a collection of tenets or opinions can be called Opinions of the physicists’ (the same is true of the alternative title ‘Physical opinions) and includes the opinions by the three main Eleatics, Parmenides, Zeno and Melissus, this means that they are supposed to contribute to physics. This assumption has some plausibility with limitation to the second part of Parmenides’ poem, but is not at all obvious in the case of the other two, who (as it is well known) offer nothing which is like a cosmogony and cosmology. It is an assumption that goes back to the interpretation that Aristotle (to some extent following Plato) offers of the Eleatics as maintaining that all things are one’, where ‘all things’ are supposed to coincide with a universe that is physical or material9. It is an assumption that can be questioned (and has been questioned) in various ways, e.g. by pointing out that Parmenides, whatever the precise subject he admits for his ‘is’, is concerned primarily with being as such, not with oneness or with immutability attributed to a (physical or material) universe.

10However, beyond the questionable character of this assumption, the fact is that much of what is stated by him or by the other two Eleatics can be fitted only by force into the sort of schematization which underlies Works like that by Aëtius. Does Parmenides admit one element or principle or more than one, e.g. two? What is matter for him? Or body? What is space for him? The situation is even worse in the case of Zeno and of Melissus, who do not provide any sort of cosmogony and cosmology. It is not difficult to notice that most testimonies concerning them do not enter at all in the category of the ‘opinions of the physicists’. Thus, in the case of Melissus, the longest and most important testimony comes from the pseudoaristotelian De Melisso Xenophane Gorgia, and includes an exposition of his thought which in the main is a paraphrase of what we get from the fragments, followed by criticism. The remaining fragments, with few exceptions, all concern general points of his doctrine (e.g. the admission that being is one and unmovable). And if one testimony depending on Aëtius (Stobaeus I, 21, 3a (3) = Aëtius II, 1, 6 = 30 A 9 DK) tells us that for Melissus the universe is infinite, but the world is finite, this looks like an attempt to attribute to him some sort of cosmology. In the case of Zeno all that we have, apart from a few testimonies about his life and work, is a collection of arguments against plurality or movement. And, in their case, we come back to the difficulty of separating fragments from the testimonies in which they are embedded, since between a Verbatim quotation, that may be incomplete, of a passage containing an argument and a close paraphrase of a certain argument of his, the difference tends to be rather small, especially since the paraphrase can include some original terminology.

  • 10 Text and translation can be found in Osborne, 1987, 328-37. It has to be admitted that Heraclitus i (...)

11And one can wonder whether some of these considerations are not to be extended, mutatis mutandis, to some other Presocratic philosopher, at least to Heraclitus, who is not the typical physicist and cosmologist and who writes in a rather aphoristic way. Some of the quotations we are given from his work are too short to be sufficiently understandable without taking into account the context of the ancient author who supplies the quotation, though some of the comments that are given in this way can be false or misleading. He is rather little represented in the testimonies that can be supposed to go back either to Aëtius or directly to Theophrastus, while the summary of his thought that is given (together with quotations) by Hippolytus contains little that could be fitted under the chapter-headings that are adopted in a collection like that by Aëtius10.

  • 11 For more on the problems raised by an edition of the Pythagoreans see Burkert 1996.

12Even the Pythagoreans, though for different reasons, do not fit very well into that sort of schematization. Some of the information we have is about the figure of Pythagoras and the events that concerned his school or sect and the practices adopted in it, and here the issue is to distinguish what is legendary from what is historical. As to the contributions by the Pythagoreans, they often belong to mathematics and to symbolism rather than to physics or cosmology, unless one wants to maintain that their contributions to physics and cosmology are imbued with mathematics and symbolism, a fact that in any case makes them rather peculiar by comparison to those thinkers that are without restriction physicists and cosmologists. Only in the case of some of them (especially of Philolaus and Archytas) fragments prevail on testimonies, but then the problem becomes which of these fragments are authentic. One can expect an edition of the texts (as is the case with the one by Huffman on Philolaus, 1993) to go beyond the limits of that by Diels and Kranz in order to clarify this point11.

  • 12 This de facto applies to the second part of the poem, e.g. by using test. 37 (Aetius II, 7, 1) to f (...)

13Talking, at this point, more directly of (real or possible) editions of Presocratic philosophers, one way of deviating from Diels’ approach lies in going much further than he does in admitting that testimonies are to be treated as complementary to the fragments, since they either constitute their context or they integrate them. This proposal seems to be applicable especially in the case of the Eleatics. It has been applied, it seems to me successfully, in the case of Parmenides in the German edition provided by Hölscher 1969, by ordering the testimonies in such a way as to (partially) fill in the information that is not obtained from the fragments, in addition to providing their context12. No doubt this procedure is successful because it is based on a previous selection of the testimonies, leaving out those that do not fit (they are in fact omitted by Hölscher). But these other testimonies could be treated as a separate block complementary to the fragments without having to suppose that they are so homogeneous with the other that they must be presented as a whole. Some of them are biographical, and these constitute a class per se in any case. Some others offer (interpretative) summaries of the thought of Parmenides, often restricted to the part which is already known to us through the fragments. They concern either general aspects of his thought (e.g. the assertion that is attributed to him that ‘all is one’ or the connection he established between truth and opinion) or more specific ones. Some of these also say something about the relationship between his thought and that of other thinkers. These certainly cannot be regarded as integrative of the fragments, in the sense that they fill gaps in our information on the contents of his poem. And, once we add to them those that do supply some information on the missing parts of the poem, very little is left.

14An edition similar to Hölscher’s is the one given by Jean Bollack of Empedocle’s poem Περὶ ϕύσεως, also published in 1969. Here again testimonies are used to fill the gaps between the passages of the poem which have been preserved, so that the interlacing between fragments and testimonies permit us to get a clearer picture than if the testimonies were kept distinct. Here again the testimonies are provided in a selective way, and a part of them (especially the biographical ones) were meant to be published in a separate volume. One cannot expect, of course, that in the case of all the Presocratics our documentation be such that fragments and testimonies can complete each other, however imperfectly.

15In the case of Parmenides Diels adopted a compromise solution, for he kept most testimonies distinct from the fragments but was induced to quote some of them contextually with the fragments. And, if one extends the attention to the other Eleatics, that is to Zeno and to Melissus, one finds that Diels respects the standard presentation he offers of the Presocratics only in form. This is certainly true in the case of Melissus, for the reason already given above. It is true in a way also in the case of Zeno, for there is a close connection between most testimonies and the fragments (they all contain arguments against plurality or against movement) and some testimonies (starting with that by Plato in his Parmenides) serve the general purpose of clarifying Zeno’s intentions in his writing (these are under the title Schrift). The connection between testimonies and fragments containing arguments is so close that Lee, in his edition (1936), abandoned the distinction between fragments and testimonies, with a result that should not induce one to suppose that it constitutes a regress by comparison to that of Diels (and it is fuller than his with regard to the evidence on the arguments, while omitting the general testimonies and the biographical ones).

16In the case of the Eleatics, however, there is not only the difficulty arising over the relationship between fragments and testimonies, which is such that (with the possible exception of Zeno) the testimonies can be treated as subordinated to the fragments. There also is the difficulty that (as already suggested above) the testimonies are often of such a nature as not to be very appropriate to give a clear and satisfactory idea of their thought. The way in which certain of their contributions is classified or presented is artificial (e.g. under the heading of those who postulate a principle that is one and immutable since they are failed physicists) And this has to do with the origin of those testimonies, in so far as they go back to a collection which is meant to assemble the opinions of the physicists’. An edition that is offered of any one of them must take into account the peculiarities of the evidence, giving up the idea of uniformity of presentation for all the Presocratics.

  • 13 Cf. Taylor 1999. The selection of the testimonies in this collection is rather restrictive (some of (...)

17In general, giving priority to fragments with respect to testimonies is satisfactory when the fragments are such, in quality and quantity, as to permit us to get an idea of what is going on in the original work from which they come or at least to have a rather full understanding of the thought of the philosopher involved in its nucleus or at least in its most important sides. This possibility is certainly excluded in the case of the Milesians, when the longest quotation we have are a few words from Anaximander. It is excluded in other cases, and one of them is precisely Democritus. One only has to look at Christopher Taylor’s edition, which, adopting the format of «the Phoenix Presocratics» series to which it belongs, keeps fragments distinct from testimonies and gives priority to the former13. By comparison with Diels’ edition the organization of the fragments is much ‘cleaner’ and more understandable, since it follows the same order of topics as the testimonies and does not include such miscellaneous material as context of fragments or testimonies about titles of works (these are replaced by reference to the testimonies), passages which should rather be taken as imitations (e.g. Diodorus Siculus I, 7-8), and quotations of single words or expressions. What remains at this point that can be said to be intelligible? Only the passages on the epistemology and on the ethics! Otherwise we have, for instance, with reference to the section entitled «Works on nature», D 7 on cosmogony («A swirl of forms of all kinds was separated off from the totality») which is followed by D 8 on sexual intercourse being called a «mild apoplexy».

18One has not to stick to the idea of uniformity of presentation even in the case of the same Presocratic thinker, for it is conceivable that different editions, for different purposes, be made of the written evidence concerning him. This, at least to some extent, has happened in the case of Heraclitus. In addition to the edition by Diels and Kranz, which of course follows the same criteria (distinction between fragments and testimonies, etc.) which are followed with the other Presocratics, there is (D the edition by Charles Kahn, 1979, which is limited to the fragments (edited in the best possible way) and which orders them in such a way as to get most meaning out of them, on the assumption that they are not aphoristic in the sense that they are to be understood in isolation from each other; (2) the edition by Miroslav Marcovich, 1978, which, while also ordering the fragments in a similar way, provides all possible variants, paraphrases, summaries and echoes or imitations for each fragment, so as to offer a complete picture of the tradition concerning them; (3) the edition (in fact only translation) by Jonathan Barnes, 1987, in his Early Greek Philosophy collection, where the context of the fragments is considered, especially by giving a continuous quotation not only of the chapter of Diogenes Laertius (this happens in Diels-Kranz too) but also of a rather long piece in Hippolytus and a less long one in Sextus Empiricus.

19This point can of course be extended to Democritus. Apart from the fact that it is not too likely that Diels-Kranz’s edition, being so convenient, will be wholly abandoned in his case (as it has not been abandoned in the case of other authors), it is conceivable that some other new edition be prepared with different purposes than mine, just as this is preceded by Luria’s edition. I can see for myself (and it has also been pointed out to me with some force by a pupil of mine who is reading the material) that my edition, with its complementary translation, is too large and complicated to be of easy use for readers who are not already familiar with the first atomists, and even they need time to get orientated. But I cannot see how to make it slimmer and more practical without abandoning some of the purposes for which I have prepared it, so that I can only try to help the readers with instruments such as indexes and concordances.

20Diels himself was not fully satisfied with the way in which the documentation concerning Democritus was organized in his Vorsokratiker, because in the preface (Vorrede) to the fourth edition he declares he could not carry on, as he wished, a revision of the work which, among other things, would be such that «das Demokritkapitel würde eine übersichtlichere Form gewonnen haben». My edition offers even less Ubersichtlichkeit, so that there is space for an edition which has this quality, while filling certain gaps in Diels-Kranz’s edition which at this point make it too inadequate for a serious study of Democritus and Leucippus.

21What can be said of my edition is that, whatever its merits and defects, it provides a certain itinerary for reading the documentation we have about Democritus and Leucippus. This itinerary is one among various possible itineraries and is influenced by my philosophical interpretation of their position. The same is certainly true also of Luria’s edition, which differs from mine, first of all, because he takes Democritus as a mathematical atomist while I take him as a physicist, but this différence is not relevant to the issue of Übersichtlichkeit. The procedure I am propounding is likely to meet the disapproval of those philologists who expect that an edition should follow criteria that are neutral enough as to leave no room for the personal preferences (even if philosophically motivated) of the editor. Yet it can be pointed out that Diels, in pursuing this aim of neutrality with the Presocratics, went as far as to organize certain fragments (in Democritus as well as in Heraclitus) by following the alphabetical order of the names of the ancient authors who are their sources, that is to say an order that is haphazard rather than truly neutral. In the case of Democritus, as in the case of Heraclitus, it is hardly possible to offer an edition of the texts that is not influenced by philosophical (and not only philosophical) interpretation, though not exactly for the same reason. In the case of Heraclitus this is so because we do not know (with some possible exceptions) the original order in which the bits left us from his work found themselves, so that we are obliged to ask ourselves which way of ordering them makes most sense. In the case of Democritus this is so for other reasons, as we shall see.

22In saying this, I do not want to claim that this procedure should be generalized, especially by being extended to all of the Presocratics. Whenever (as in the case of Parmenides and of Empedocles) it is a matter of trying to reconstruct from fragments the original sequence of propositions of a given lost work it is of course mandatory to collect all possible evidence which shows what the actual position was of those passages in that work. There is space for the personal choices of the editor only when this information (as often happens) is insufficient.

II. Sources and traditions

23Democritus is to some extent a peculiar case even among the Presocratics, for the fragments that are preserved (questions of authenticity aside) are numerous, but nearly all belonging to his ethics, whereas, if we leave these out, fragments are so few that in the case of nearly all his doctrines we are obliged to rely exclusively on testimonies. Testimonies, again, are numerous, but in many cases they give the same information with only slight differences in formulation. What remains true is that he, in contradistinction with the Eleatics and even to Heraclitus, can be treated as a ‘physicist’, and as a physicist with such a variety of interests as to have contributions to offer on almost any of the topics which are suggested by the titles of the chapters of works like that of Aëtius. Considering him this way does not cover all his contributions, even leaving out his ethics, but it certainly covers the most important and the most influential part of them.

  • 14 Indeed there is only one sure reference (Taurus apud Philoponus, De aeternitate mundi contra Proclu (...)

24As with the other Presocratics, whose main contributions are in the field of physics (in the ancient sense of the word) and cosmology, one question that arises is how far the testimonies for them belong to one tradition, in that (as Diels assumed) the sources of testimonies in fact depend (rather exclusively) on one work, identified with Theophrastus’ collection of ‘Opinions of the physicists’, which thus should be regarded as the Urquelle. This question leads to questions that have already been raised by other scholars: given that Simplicius and other ancient authors sometimes mention Theophrastus’ Physics as their source and that their references to that collection are not frequent14, and given that we cannot be certain that the De sensibus constitutes one book of that collection, are we sure that there is just one work by Theophrastus’ that is the point of departure of all the other collections of that sort? And given that Theophrastus adopts an approach which is very close to that by Aristotle, how can we exclude that in some cases those collections go back directly to Aristotle? Other questions than can be raised are about the linearity of a tradition which is supposed to be reducible to two main collections, one of which is the abridgement (επιτομή) of Theophrastus’ work, i.e. the so-called Vetusta placita, and the other (that by Aëtius) is the epitome of the latter work.

  • 15 From now on I add the abbreviation «Le. » (to be kept distinct from «Lu. » for «Luria») when giving (...)
  • 16 On the parallels between Aristotle’s treatment of the earth in De caelo II, 13 and Ps.-Plutarch’s E (...)

25Concerning Aristotle as a source, Democritus constitutes an interesting case, which deserves some attention. It is well-known that Simplicius, who is one of our main sources of information about Democritus, was acquainted with Aristotle’s lost monograph on Democritus, since he quotes an ample passage from this work (cfr. 68 A 37 DK = 7.1 + 79.4 Le.15) In other cases in which he, in commenting some Aristotelian passage, either makes it clear that it refers to Democritus or/and expands it by adding some information about the thought of the atomist, one gets the impression that he is always depending on Aristotle, though he may be using passages of the works we possess rather than that lost monograph on Democritus, and the same must be true of other commentators like Philoponus. What is sufficiently evident in the case of the commentators could have happened with some doxographers, perhaps including Aëtius, whose way of working is unknown to us16.

  • 17 A reference to the monograph in two books on Anaxagoras is made by Simplicius in his comm. on Phys.(...)
  • 18 The title Φυσικά (excluding φυσικοί as a possible reading of the genitive, see below) is used by Al (...)
  • 19 As Diels assumed, taking it, in a questionable way, as equivalent to the Περὶ φυσικών ἐπιτομή in 2 (...)
  • 20 Cf. fr. 5a Diels (= 231 FHSG).
  • 21 This is how Mansfeld and Runia, 1997, 183, n. 203 (with reference to other contributions by Mansfel (...)

26This last comment goes beyond the case of Democritus, while still concerning Aristotle’s role. But even Theophrastus’ role does not seem reducible to the influence exercised by his collection of ‘opinions of the physicists’. In the first place it should not be overlooked that also Theophrastus’ monographs on single thinkers could have been exploited17, as this is parallel to the use of Aristotle’s monograph on Democritus, but no doubt these monographs could reflect essentially the same approach as that collection. In the second place, it is very likely that also Theophrastus’ Physics was used as a source, since this work is explicitly mentioned by Alexander, by Simplicius, by Diogenes Laertius and possibly by Aëtius as a source for information of doxographical type18, with the comment (in the passage by Diogenes Laertius) that there he set forth the tenets of practically everybody. In the third place, Diogenes Laertius (in IX, 21 = fr. 6a) mentions as a source of the information he gives about Parmenides an Epitomé by Theophrastus which must be the same as the ‘Επιτομὴ ϕυσικῶν δoξῶν, if a distinct work with this title19 is what Galen mentions (using the plural) in his commentary to the De natura hominis attributed to Hippocrates20, rather than simply suggesting that Theophrastus’ main collection provides summaries of the doctrines of thinkers like the same Parmenides or Xenophanes21. In the former case this work could constitute an important point of departure of other Epitomai like that by Aëtius, which thus could be based on a previous work similar to them rather than being the result of an ‘epitomizing’, i.e. an abridging (by synthesis and omission of information), of a larger work like the Physicorum opiniones, which according to the catalogue of Theophrastus’ works was in 16 books (cfr. Diogenes Laertius, V, 48). I leave this question open.

  • 22 Cf. his commentary to Aristotle’s Physics, 115.11f. Diels, concerning Parmenides (= fr. 7 = 234 FHS (...)
  • 23 This analogy was already propounded by Diels 1879, 102.
  • 24 This is normally quoted by him as Φυσικά e.g. at the beginning of his commentary to Aristotle’s Phy (...)
  • 25 This is the position that was already suggested by Gottschalk 1967, is contemplated by Sharples 199 (...)

27How far Theophrastus’ Physics should be regarded as an important indipendent source to the detriment of the Physicorum opiniones is a matter of controversy. If we admit that Diogenes Laertius IX, 22 is reliable in the aforementioned comment, and that the title Φυσικὴ ἱστορία that is adopted by Simplicius in three passages which offer information of the doxographical type22 is to be understood as equivalent to Περὶ ϕύσεως ιστορία (= Research concerning nature) and thus as an alternative title for Φυσικά, its importance becomes such as to overshadow the other work, which would be quoted only by one source apart from its appearance as a title in the catalogue of Theophrastus’ works (cfr. above, n. 14). If, on the other hand, we suppose that Φυσική ιστορία is to be understood as meaning History of physics (on the analogy with such Peripatetic works as Eudemus’Γεωμετρική ιστορία)23 and as being the title by which Simplicius refers to the Physicorum opiniones, since he apparently treats it as a work distinct from Theophrastus’ Physics24, the order of importance between the two works is inverted25. However attractive this second suggestion may be, I think one has to suspend one’s judgement, for it has the disadvantages of obliging us to regard Diogenes Laertius’ testimony as unreliable and to regard the fact that the only work referred to by Simplicius before giving his main testimony from Theophrastus is the Physics as having no weight.

  • 26 Hence Kühn 1828, 25, translates as follows: «... in quibus naturales opiniones in compendium redegi (...)
  • 27 Cf. Mansfeld 1992, 64f., and n. 10. Also Sharples 1998, 10, points out that it would have been easy (...)
  • 28 This has been pointed out by Zhmud 2001, who adduces evidence. (In what follows I stick to the conv (...)

28Even the title of Theophrastus’ collection of ‘opinions of the physicists’ is a matter of controversy. However we interpret Galen’s testimony, it would seem that he understood Theophrastus’ work as epitomizing physical opinions’, since, in drawing a parallel with Meno’s «History of medicine», he could not maintain the parallel so close as to say that, just as that one offers an account of opinions of ancient physicians (δόξαι τῶν παλαιῶν ιατρῶν), this offers an account of their (= the physicists) opinions (ἱστορία περὶ αὐτῶν δοξῶν), but said that it offers an account of these (= opinions rather than persons: ἱστορία περί τούτων)26. In any case the most obvious translation of ‘Επιτομὴ ϕυσικῶν δοξῶν (if it is a title) is Epitome of physical opinions, just as the most obvious translation of Περί τῶν ϕυσικῶν δοξῶν (as the title is given in its only full occurrence, apart from that in the catalogue of Diogenes Laertius, V,48, where it is equally in the genitive) is, as Mansfeld points out27, On the physical opinions, so that it would seem that this (rather than On the opinions of the physicists) is likely to be the title of Theophrastus’ main ‘doxographical’ work. On the other hand Aristotle’s commentators usually talk of ϕυσικῶν δόξαι and not of ϕυσικαι δόξαι28. In any case the accent in the title remains on the opinions, not on the philosophers, and even Diels admitted that the work, in its main lay-out, was organized by topics, not by philosophers (or groups or schools of them), while to some extent following a chronological order inside each book. But this makes it difficult to adopt Diels’ suggestion (1879, 102) that the title in Diogenes Laertius V, 46: Περὶ τῶν ϕυσικῶν in 18 books, is just an alternative title for the same work (evidently taking it as meaning about the physicists’ rather than ‘about physical things), even if we overlook the difference in the number of books, for this would definitely concern philosophers and so not be organized by topics. (Hence also the work referred to by Alexander in the passage mentioned in n. 18 is not about the physicists). It would also be odd that the same work appeared under different titles in a catalogue, though it is surprising that a big opus in 18 books should have disappeared without leaving any trace behind. On the whole I conclude that it is on this point, rather than on the other, lesser point, that Diels’ position is open to criticism.

29As to the linearity of a tradition depending, if not exclusively on one work by Theophrastus, on his works and on those by Aristotle, there is Mansfeld’s suggestion (in criticism of Diels) that the collections of placita underwent frequent changes by additions or by abridgment or in other ways and that they belonged to more than one tradition.

  • 29 Mansfeld 1996, 23f. (also cfr. Mansfeld-Runia 1997, xix-xx).

«Diels also failed to take the possibility into account, at least not in a consistent way, that already before Aëtius more than a single tradition existed, or mutually diverging witnesses belonging to the same tradition may have been available (just as, though to a far smaller extent, they are available today). Furthermore, those who contributed to the tradition (like the tellers of tales, or the performers of epic song in an oral tradition) were in no way obliged to preserve the material found in their predecessors unchanged. Handbook literature serving a practical purpose has to be updated and revised, and will mirror the needs and fashions of the time. »29

30It is not difficult to realize that these observations are relevant for Democritus. A significant part of the testimonies we have introduce a comparison between his position and that of Epicurus, so that, for obvious chronological reasons, they cannot go back to Theophrastus. As is natural, doxographers and epitomators like Aëtius were induced not simply to abridge but to make additions, taking into account recent developments in the field of physical speculation. What is more difficult is to say how these additions were made.

  • 30 Cf. Stobaeus I, 14, 1 f(2-3) = 68 A 47 = 64.7 Le. ; Ps.-Plutarch, Placita I, 3, 9 = 68 A 47 = 9.1 L (...)
  • 31 E.g. that he introduced the clinamen as an uncaused deviation of the atoms from their course to avo (...)
  • 32 Not made by Democritus. (It is likely that Epicurus himself is rejecting the determinism he attribu (...)
  • 33 On this example cf. O’Brien 1982.

31Part of the testimonies to which I am referring simply State that there is some difference between the position of Democritus and that of Epicurus (e.g. that only the second admitted weight as a cause of the movement of atoms30) Another part of these testimonies is polemical in tone and formulation, in that they tend to suggest that Epicurus derived the atomic theory from Democritus (so that his claims to originality are ungrounded), but that he introduced changes in it which were to the worse31. However, if we go beyond the polemics, we find that what is stated about Democritus’ position coïncides with what the Epicureans thought his position was. Thus Diogenes of Oenoanda, fr. 54 [= 68 A 50 DK = 76.1 Le.], propounds as Epicurus’ discovery32 of a free motion of the atoms what for Cicero is his postulation of their uncaused motion, implying that this is incompatible with Democritean determinism. A less clear example is that of the greatness of the atoms, where an hypothetical consequence drawn by Epicurus from the admission of an infinite variety of figures could have been taken as representing the position adopted by Democritus33. Hence it is not unlikely that the doxographers and even the Aristotelian commentators depend on some Epicurean source for this sort of account of the relationship between Democritus and Epicurus, though it is impossible to say whether this means there was an Epicurean tradition of doxography (in the narrow sense of the word) on which the Aëtian one depends.

  • 34 This only happens in some of his letters of which fragments have remained (2 based on Philodemus, c (...)
  • 35 Cf. Epistula ad Pythoclem §§ 89-90 (= 67 A 24 DK = 81.2 and 80.3 Le.) on cosmology; Epistula ad Her (...)
  • 36 On the composition of the soul, cf. III, 370-3 (= 68 A 108 DK = 103.6 Le.)
  • 37 Cf. fr. 6, coll. II 9-14, III 1 (= 5.12 Le., om. DK), fr. 7, col. II 2-14, III 1-2 (= 8.4, om. DK) (...)
  • 38 Perhaps it is not a chance that Diogenes Laertius’ work concludes with Epicurean doctrine and that (...)

32Of course the Epicureans are our source of information about Democritus (and Leucippus) not only in this indirect way but also directly, but this is true with certain restrictions: Epicurus himself, in the Works we possess, never mentions Democritus by name34, but appears to make polemical allusions to him a few times35; Lucretius makes reference to him on some rather specific doctrines, only in one case expressing dissent36; Philodemus’ references, however interesting, have little or nothing to do with atomism; there only remains Diogenes of Oenoanda who takes position in a critical way on important issues of atomistic theory37. In addition we have some knowledge, in an indirect way (i.e. through Plutarch), of Colotes’ criticism of Democritus in a work of his meant to show that all philosophers (except of course the Epicureans), with their theories, make life impossible. This same motive, with application to Democritus, appears in a fragment by Diogenes of Oenoanda (cfr. fr. 7 Smith = 8.4 Le.), so that it is possible that he depends on Colotes. If we add to this that we are informed that Philodemus dealt in an ample way with the thought of the main philosophers (or philosophical schools different from Epicureanism), there is enough evidence to show that the Epicureans dealt with the thought of their predecessors and that this constitutes a tradition that is at least in part independent from the tradition of doxography that has its origin in the Peripatetic school38. It certainly is different from this tradition in that it is centered on single philosophers or on schools, not on topics. But this difference in approach need not exclude some overlapping in contents (e.g. there is the suspicion that they exploited a skeptical exposition and interpretation of the thought of a certain philosopher to condemn this thought itself as infected by skepticism).

  • 39 I differ from Diels-whose quotation (68A112) starts at 1009b7-and from other scholars in admitting (...)

33Another tradition can be recognized in a series of testimonies concerning either Democritus’ epistemology or his (connected) distinction between appearance and reality (often using as a point of departure his well known assertion that ‘by convention colour etc., in reality the void and the atoms). This is constituted by a substantial part of Sextus Empiricus’ and of Galen’s testimonies concerning Democritus, by three testimonies in Cicero’s Academica, by one testimony in Diogenes Laertius (Lives IX, 72), and possibly also by some testimonies due to Plutarch. There is an overlapping between these testimonies and those by Aristotle in Metaphysics IV, 5 (especially 1009a38-bl639) and by Theophrastus in some passages of his De sensibus (esp. §§ 63-64 and 69-72), but the approach is clearly too different to permit the supposition that the aforementioned series of testimonies depend on those by the Peripatetics. Aristotle includes Democritus in a sort of genealogy (initially suggested by Plato in his Theatetus) of philosophers who were sensists, i.e. tended to assimilate thought (hence knowledge) to sense-perception. Theophrastus, apart from the mentioned agreements with Aristotle’s testimony, is concerned with Democritus’ account of the mechanism of sense-perception and of the constitution of perceptible properties. Little or nothing of all this is to be found in those post-Theophrastean testimonies, which rather stress the devaluation of sense-perception as the source of knowledge.

  • 40 That Arcesilaus used to appeal to the authority of other philosophers-though of these only Socrates (...)
  • 41 Because the quotations he gives are rather short, when they are given at all, and in Adv. math. VII (...)

34As to the source of those testimonies, it is easy to notice that the passage in Diogenes Laertius (IX, 72 [= 68 B 117 = 61.1 Le.]) appears in the chapter on Pyrrho and belongs to a context in which Democritus is associated to Xenophanes, Zeno of Elea, and still others, as a precursor of Pyrrhonian skepticism. Quite similar to it are two of the three testimonies in Cicero’s Academica (i.e. II, 23, 73 [= 68 B 165 = 61.2 Le], and I, 12, 44 [= 59 A 95 = 61.4 Le.]), for there too Democritus is mentioned with others (including Socrates, Anaxagoras and Empedocles) as a precursor of skepticism. Only that this time the skepticism that is considered is that of the New Academy, with reference to Arcesilaus as its founder40. Sextus’ position is somewhat different from those attributed to Pyrrho and to Arcesilaus, since he wants to underline the originality of Pyrrhonian skepticism by pointing out the differences from other positions, including not only that of Democritus (cfr. esp. Pyrrhoniae hypotyposes I, 30, §§ 213-14 [= 58.1 Le., om. DK]) but that of the New Academy (cfr. op. cit., I, ch. 33), and does so in apparent polemics with those other skeptics who underlined continuity. This however does not exclude that Sextus himself depended on the same kind of source of information as these other authors, unless one wants to maintain (as I am not inclined to do41) that he had before him original works by Democritus.

  • 42 These are the two passages from De medica experientia: IX,5 [p. 99 Walzer = Nachtrag in vol. II DK, (...)
  • 43 Because certain misunderstandings of his on Democritus’ position show lack of information, but this (...)
  • 44 This conclusion, not extended to Galen and to Plutarch, was already drawn by Decleva Caizzi 1980, a (...)

35It is difficult to go beyond the formula just used by me («same kind of source of information»), as various possibilities remain open, of which the use of (precisely) the same handbook is only one. Still, there is sufficient continuity internal to this tradition, and sufficient difference from any other tradition, to make us admit that there is a tradition which is distinguished by the fact that it is connected with skepticism in both the Pyrrhonian and the New Academic version. And it seems rather likely that the testimonies by Galen, which (in two cases) are supposed to reflect the position of the Empirical physicians, i.e. a position that was close to skepticism, belong to the same tradition42. His other main testimony (from De elementis ex Hippocrate I, 2 = 68 A 49 = 8.3 Le.) is too close to what we find in Sextus to have to admit another sort of source. Whether certain of the testimonies by Plutarch, which deal with the same kind of topics, belong to the same tradition is more difficult to tell, for there are points of contact but also points of difference, and in his case too it has been claimed he had original works of Democritus before his eyes, but I do not think this can be so43. Anyhow, whether we can include him or not, the general conclusion seems to be sufficiently established44.

36This group of testimonies constitutes another difficulty (after what has been said above concerning mainly the Eleatics) for the assumption that we can bring all Presocratic thought under the heading of physics and on this basis make it depend on one ultimate source, the exposition of physical opinions due to Theophrastus, for epistemology is not usually considered as belonging to physics (this is clear with the Stoics, who bring treatment of criteria of truth under logic, but I think it can be shown that Aristotle’s attitude on this issue was not much different). But the same is true, in different ways, for other contributions by Democritus about which we get information in the testimonies we have. On the whole it is difficult, apart from the cases already considered, to associate groups of testimonies with sufficiently distinct and continuous traditions, except perhaps those non-authentic depending on Bolus. But it is clear enough that a significant part of the testimonies concerning Democritus do not depend on the tradition that goes back to the Peripatetic account of the opinions of the physicists.

  • 45 I cannot believe that Democritus wrote a book on Pythagoras, of which otherwise we know nothing, an (...)

37In the first place, the classificatory and introductory work (to the books of Democritus) done by Thrasyllus must be of this kind, for it must depend on similar classificatory work done by Callimachus and it reflects the Pythagorean point of view of the same Thrasyllus (it is typical of this point of view that he puts the title «Pythagoras» at the head of the list of Democritean books45).

38In the second place, as it has already been pointed out by some scholars who have dealt with them (e.g. Rütten 1992, Smith 1990), some of the pseudo-Hippocratean letters betray some knowledge of certain titles of Democritean works and of certain of his doctrines (e.g. the existence of infinite worlds). But nothing suggests that the source of information of the author of the letters coincides with that of any of the other authors who belong to any of the traditions mentioned so far.

  • 46 Cf. Nat anim. VI, 60 (= 68A150a DK = 96.8 Le.), XII, 16 (= 68A151 = 96.7 Le.), XII, 17 (= 68A152 = (...)
  • 47 Some of the sources of Aelian have been studied by M. Wellmann in articles that appeared in Hennes (...)

39In the third place, there are some testimonies on specific physical issues which are significant, because they are rather detailed or for some other reason. Among them there are two by Seneca on winds and earthquakes (cfr. Naturales quaestiones V, 2 = 68A93a = 87.1 Le., and VI, 20 = 68A98 = 87.3 Le.), one by Diodorus Siculus on the flood of the Nilus (cfr. Bibliotheca historica I, 39, 1-4, referred to in 68A99 = 88.2 Le.), one by Alexander of Aphrodisias on the (lode-)stone which draws iron (cfr. Quaestiones II, 23 = 68A165 = 89.1 Le.), and no less than 7 passages by Aelian on facts concerning animals46. It is difficult to say what origin these have. Perhaps some of them are survivals of the original exposition by Theophrastus which has not been replaced, as in other cases, by the shorter versions contained in some epitomé (an indication in this sense is given by the fact that there are shorter versions, one of them in Ps.Plutarch’s Epitomé [cfr. IV, 1-4 = 68A99 = 88.1 Le.], of Democritus’ account of the flood of the Nile). However this cannot be true of all, certainly not of the passages by Aelian, for a collection like that by Theophrastus (if we can reconstruct it from «Aëtius») had no space for certain details on animals. Only a close study of the sources of these authors can perhaps throw some light on this point47.

  • 48 Cf. Runia 1999.

40There is, furthermore, a not small number of testimonies which give us an idea of the variety of Democritus’ interests, which can be called encyclopedic and which suggested in antiquity the comparison of him to a pentathlet (cfr. Diogenes Laertius IX, 37 and Suda, s.v. pentathlos). These interests cover, among other things, Homeric criticism, grammatical questions, etymology, geography, medicine and agriculture. The sources for all this information are too varied for us to conjecture any unitary tradition in their transmission: they are independent not only from the Theophrastean tradition but also reciprocally. (Aëtius of course included statements by physicians48, but these concern physiology and anatomy, not strictly medical questions.) Finally (but my survey is not meant to be complete) there are testimonies, by Plinius and others, on Democritus’ alleged contributions to magic and alchemy. There are sufficient reasons (because what Plinius asserts is refuted by Aulus Gellius and by Columella, and the latter attributes one work, i.e. the Cheirokmeta, regarded as authentic by Plinius, to Bolus) to admit (with Wellmann and Diels) that these testimonies, which gave rise to the legend of Democritus as a magician, actually concern, at least in part, Bolus. It may also be significant that, while there are a number of biographical testimonies which tell tales about Democritus not different in kind from those which are told about other Presocratics like Heraclitus, these do not suggest in any way this image of Democritus, while some authors (such as Plutarch and Lucian) stress his rationality in his search for physical causes.

41From the survey I have just provided, however incomplete, it is sufficiently clear that the testimonies concerning Democritus cannot be regarded as wholly or almost wholly depending, directly or indirectly, on Theophrastus’ collection of the opinions of the physicists. And this becomes one reason why ordering these testimonies by following the order of the topics adopted in that collection (or in successive collections that depend on it) is not very satisfactory. Diels himself was able to largely respect that order because he put under heading B (fragments) quite a number of passages, such as some of the epistemological passages (a few of them were completely omitted) and a large part of the ‘encyclopédic’ ones, while those he regarded as not authentic (including some, esp. medical ones, which in my view do not deserve such skepticism) were collected under a separate section of «Unechte Fragmente». But this way of treating as fragments what in most cases are testimonies (at most including some quotation) is not satisfactory either.

42Even if one had not to resort to some artificial solution like this one to preserve the Theophrastean order, supposing that this is rather faithfully maintained in later collections, the question of whether it is justified to apply it to Democritus would still remain. (This question can be raised with other thinkers, beyond the difficulty previously discussed of treating some of them as physicists’, but here I concentrate my attention on him.) I doubt that so much faithfulness is maintained, since it does not seem likely to me that Theophrastus himself dedicated an important part of his Physical opinions to contributions to «theology» (when this is kept separated from cosmology and cosmogony), but, whether he did or not, it has to be remarked that there is some difficulty in identifying some sort of ‘theology’ in the case of the Presocratics in general. In the particular case of Democritus, Diels supposes his contributions to it are attested by either those passages (by Sextus and Plutarch) that mention εἴδωλα or by those (by Cicero etc.) which assimilate God to the intellect. And these passages are put after a series of passages which concern not only the principles or elements but other topics such as space, void, time, movement, necessity and chance (cfr. 68 A 74-79 + B 166). Apart from this deviation from the order he professes to respect, it is more natural to associate these passages either to the treatment of the soul (as I do in my collection) or (with limitation to the first group) to the treatment of the εἴδωλα, as happens to some extent among the Epicureans (cfr. Lucretius IV, w. 129-142), unless one were to restrict the whole topic to the issue of the origin of religious belief (dealt with by Lucretius in V, w. 1161 ff.).

43This example illustrates the difficulties one meets in applying to a Presocratic like Democritus the order that is adopted in collections like Pseudo-Plutarch’s Placita. This order, whether it does go back to Theophrastus himself or not, is only to some extent an order that was largely adopted by ancient authors, and for the rest is either a typically Peripatetic order or an order which is somewhat arbitrary. Typically Peripatetic is an order which adopts a clear distinction between meteorological (including geophysical) and celestial issues (in Ps.-Plutarch’s Placita the former are dealt in book III, the latter in book II, after some chapters on cosmology and cosmogony). Peripatetic procedure too is to keep distinct from the treatment of the soul (to be found in book IV of the Placita) the treatment of certain topics (such as the origin of sleeping and death and of dreams, to be found in book V of the Placita) which manifestly correspond to the contents of Aristotle’s Parva naturalia (as a work distinct from his De anima, though connected with it as a sort of appendix). On the other hand placing issues concerning physiology (in book V of the Placita) after issues concerning psychology is rather arbitrary. And there are further complications, and discrepancies from Stobaeus’ collections, to be discussed below, which suggest that the order adopted by Theophrastus, if it was typically Peripatetic, was not respected with sufficient consistency.

44Why should we respect this order, which certainly was not one that, on some points at least, was followed by Democritus himself? The distinction within psychological topics is too peculiarly Peripatetic, while the distinction between meteorological and celestial phenomena, though widespread, is one that starts with Aristotle, for it was ignored by the Presocratics, and is not always adopted even after him, e.g. it is not constantly adopted by the Epicureans (it is adopted by Lucretius but not by Epicurus himself in his Epistula ad Pythoclem). Still, since we are not in the position to reconstruct the order that was adopted by Democritus, it seems reasonable to follow the order that is found in a collection like Ps.-Plutarch’s Epitomé in so far as this coincides with the order that is common to all the main physical Systems in antiquity. At least the general succession: principles, cosmology and cosmogony, astronomic System, meteorological phenomena, phenomena concerning life (or the soul), is rather convenient for ordering part of the documentation.

  • 49 One can compare Stob. IV, 6, 19 (= 68B267 = 156.1 Le.), III, 1, 45 (= 68B47 = 156.3 Le.), II, 9, 3 (...)

45Once we go into further detail, however, all the reservations stated above come into play. Diels himself must have had the preoccupation of avoiding the adoption of an order which reflects the interpretative point of view of the modem scholar. But it would seem that the only way of adopting a wholly neutral order is the one followed by him for the fragments, in quoting them according to the alphabetical order of the ancient author who is the source. And this is evidently a very arbitrary order, which in various cases has the effect of separating passages which present some recognizable connection (thus the two main groups of ethical fragments are kept separate). Furthermore in some cases, as when the same passage (fragment) is quoted by different authors, this leads to difficulties (they are of course much more frequent with Heraclitus than with Democritus, since the fragments, apart from the ethical ones, for which there are coincidences in ‘Democrates’ and in Stobaeus, are few). And it is not even an order which is followed systematically, for, apart from some other smaller exceptions, the ethical fragments in Stobaeus respect the order of the topics adopted in that anthology. (I do not intend to discuss in any detail questions concerning these fragments, but it has to be pointed out that the way in which some of the fragments are put under the headings suggested by the titles in Stobaeus is often forced or arbitrary49. The same happens with other authors, e.g. a passage from Plutarch’s work on tranquillity [εὐθυμία] concerning just this notion is put under the title Περί ϕιλοπονίας presumably because ἀπονία is rejected in it!)

46It is certainly true that an order based on content, in as much as it reflects the interpretative point of view of the modem scholar, can influence the reader and incline him to adopt that interpretation. Yet, if one shows clearly, as I have tried to do by means of a detailed table of contents and of a summary of the interpretations I adopt, what this point of view is, at least the sophisticated reader, who has the patience to get acquainted with this information, is not prevented from following a path that can be quite different from that followed by the editor of the texts.

47An order of exposition which is suggested only by the subject-matter dealt with by the author cannot cover all the evidence, since those who are our source of information about his position are often influenced by interests that are different from the simple exposition of that position. This is true more or less in the case of all the Presocratics (and of the Stoics as well), but it is certainly true to an important degree in the case of Democritus. Thus Aristotle, when he discusses Democritus’ position, is often induced to collocate it among other various positions (a) either by using some unifying criterion of classification or (b) by admitting it belongs to some continuous development of thought concerning some fundamental problem. Case (a) is exemplified by the classification of principles at the beginning of Physics I, 2, according to which Democritus is supposed to admit principles that are infinitely many and in movement, and is equally exemplified by the distinction of types of infinity emerging in passages such as Physics I, 4, 187b7-12 (= 37.1 Le, to be seen with the comment of Simplicius, In Physica 165.8-18 [= 37.2 Le.]), according to which Democritus is supposed to admit infinity in quantity not only for the number of the constituents of matter and of other entities as well (certainly the worlds are included) but also for the extension of matter and of void, and is equally supposed to admit infinity in quality (or in form) because he admits an infinite variety of figures in the atoms. Case (b) is exemplified by the admission (in book I of the Metaphysics and elsewhere) of a gradual discovery of the 4 (types of) causes’ on part of the Presocratics, according to which Democritus had some grasp of the formal cause in addition to the material one, but ignored the efficient and the final ones; it is also exemplified by putting Democritus at the end of a sort of genealogy of thinkers who assimilated thinking to sense-perception (this is to be found in Metaphysics IV, 5, and is manifestly influenced by the similar genealogy propounded by Plato in his Theatetus, where however, as elsewhere in Plato, Democritus is not mentioned). Furthermore, another operation accomplished by Aristotle is to make comparisons between the position of the first atomists and some other position, by noticing similarities and differences between them. Thus he establishes a connection between the atomists and the Eleatics because the former, though propounding a substantially different position, accept some postulates of the latter, but also establishes a connection between the atomists and Plato, because they both propound some sort of atomism, theirs being physical and his geometrical. (One can notice that in first case it is not simply a matter of making a comparison but also of propounding a genealogy.)

48Considering that the approach adopted by Aristotle has a great influence not only on his commentators but on the whole doxography (whether directly or through Theophrastus), it is sufficiently clear that it is very difficult to order the documentation in a satisfactory way without taking it into account. But the order that we obtain in this way is in most cases different from the order that we get from the strict point of view of the contents of the documentation, both when it is a matter of expounding propositions as belonging to a certain philosophical System or of expounding them in the way most convenient to introduce the reader to the subject. The consequence of all this is that the adoption of an order determined by content only apparently gives uniformity to the evidence by following one criterion, because what is really done is to apply different criteria to different groups of texts. And this makes orientation difficult for the reader (as already remarked above, little Übersichtlichkeit is saved).

49In all these choices the judgement of the interpreter plays some role, which is why it can be said (as I said above) that a collection of texts of Democritus like the one put together by me offers to the reader a certain itinerary dictated by the interpreter. But this is also true, and probably to a larger extent, because a number of choices that are made are not in any way suggested by the evidence, but depend exclusively on the judgement of the interpreter on how the position expounded is to be understood or on the more convenient way to expound it.

III. The doxographical documentation

  • 50 Luria omits one of them, except for a reference to it, while Diels omits them all, except for a quo (...)

50What has been said so far leads to the question of how to handle the documentation which is more stricly of the doxographical type, in the sense that it goes back to a collection like Aëtius’ Placita or (at least ultimately and in part) to Theophrastus’ Physicorum opiniones. In the case of this documentation we often meet passages which can be considered as variants of other passages either because they depend on a common source or because one of them depends on the other. In my collection I adopt the following procedure: in the former case I give the passages in full (in my edition, with shortenings in my translation), in the latter case I give the passages in full only when there are significant differences, otherwise I give the data of the identical (or nearly identical) passage. The most obvious instances for these situations are precisely those passages, in Ps.-Plutarch’s Epitomé, in Stobaeus and sometimes in Theodoret, which depend on Aëtius and those passages, in Ps.-Galenus’ Historia philosopha etc., which depend on the same Epitomé (I assume the validity of Diels’ hypothesis as revised by Mansfeld and Runia in theirs Aetiana). Other examples can be given, one of which is the following: there are three passages by Sextus Empiricus (i.e. Adversus mathematicos X, 310-311 + 318; IX, 359-364; Pyrrhoniae hypotyposes III, 31-33) on the principles or elements of the physicists (explicitly including Democritus) which, though belonging to different contexts, are similar and may have some common source different from Aëtius: they are all to be made available in a collection on Democritus (cfr. my 6.2, 6.3, 6.4), since none of them overshadow the others so as to make their quotation useless50.

  • 51 I can only give one example (other examples in the introd. to my collection): among the passages wh (...)

51Concerning the passages depending on Aëtius, in Diels-Kranz and in most other editions and translations it is usual to quote only one passage, selected from one of the mentioned sources, under Aëtius’ name, and without further specifications. This is not satisfactory, apart from the fact that it amounts to treating a hypothesis as a fact, since in some cases there are discrepancies of some weight betwen these variants51. Another reason why this procedure is not satisfactory is that it works on the assumption that we can reconstruct Aëtius’ collection, at least to a large extent, and this is an assumption that is questionable. This assumption is connected with the wider assumption that we can trace the tradition of doxography that goes back to Theophrastus’ Physical opinions, but, as we have seen, it is not likely that what is involved is a single continuous tradition going back to just one work taken as constituting the one Urquelle. Though there is a connection between the two assumptions, the first could still be defended while the other is abandoned. Yet, before dealing with this issue, there is a point to be made that concerns both assumptions.

52This is that a weakness of the whole Quellenforschung as conducted by Diels is that he does not take into account the method (or methods) that is applied in works of doxography like that Epitome of Ps.-Plutarch. The weakness has been stressed by Jaap Mansfeld and by his collaborators (especially by David Runia) in a number of their contributions. Their suggestion is that the organization or lay-out and structure of these Works is best understood when one sees that they involve the application of a method that can be called ‘dialectical’ and that goes back to the methodical formulations that can be found in Aristotle’s Topics, Posterior analytics, etc. The method can be called dialectical because what is involved is the collection of opinions or tenets (δόξαι) in a rather systematic way, so as to facilitate their use for the purposes of discussion and research. Recognizing this fact is of help not only in tracing the traditions that are behind those works of doxography but also in reconstructing with more precision the contents of a common source, when this is available, as in the case of Aëtius for the Epitome of Ps.-Plutarch, for certain corresponding sections in Stobaeus, and so forth. Thus certain suggestions that were made by Diels in his reconstruction of Aëtius’ Placita can be corrected and improved upon.

53It has to be admitted, in my view, that this consideration of the method that is applied constitutes a significant complement to the more traditional Quellenforschung that was conducted by Diels and that in various cases it can lead to more satisfactory results, apart from constituting a useful instrument of analysis. Recognizing its importance however is not the same as accepting that this complement is sufficient to solve all the difficulties which arise in conducting Quellenforschung. Precisely in the case of the reconstruction of Aëtius’ Placita we meet difficulties, as I shall illustrate below, that cannot be overcome by an examination of the method applied.

  • 52 One cannot take for granted that he agrees with Plato in this. In fact he discusses this procedure (...)

54There are three general issues that are involved, two of which I only can touch upon very briefly, since their discussion would require a paper by itself. The first is that the account of the ‘dialectical’ method in its application to doxography which is given by Mansfeld and his collaborators is too simple to be adequate for their purposes. We should talk, rather than of one method, of a plurality of techniques that are in a complex interrelationship and which can be used with different aims. For instance, they tend to admit a rather close connection between the use of the classificatory and definitory technique of division (διαίρεσις) and the pointing out of an opposition between opinions on a given issue (what in Greek is called διαϕωνία). Yet the first technique can be used, and is often used by Aristotle (e.g. in his zoological Works), quite independently from the second, precisely for exclusively classificatory and definitory aims, and one can even doubt that he would have regarded it as properly dialectical52. The second technique is used both by Aristotle and by the skeptics, but in different ways, for the skeptics suppose that a genuine diaphônia, when detected, is not eliminable, so that one has to suspend judgement, whereas Aristotle looks for a third, superior, position (the one propounded by him) that overcomes the unilaterality that is presented by the opposing positions that give rise to the diaphônia. (E.g. the third position between denying the existence of the infinite and admitting its existence is to admit that is exists only potentially). This Aristotelian use of diaphônia leaves room for criticism of each of the opposing positions, while this tends to be excluded by its skeptical use.

  • 53 One case is that discussed by Mansfeld himself, 1990 (p. 3097), of the location of the ruling part (...)

55I try to give a brief illustration of how these remarks have application to doxography. According to Mansfeld 1996, «the structure of numerous Aëtian chapters is that of a diairesis, with the emphasis on disagreement, while odd tenets, or tenets which do not easily fit the general structure, are listed at the end. Accordingly the tenets are more important than the name-labels affixed to them, and so these name-labels more often than not are cited in a systematic not a chronological sequence-as already in Theophrastus’ De sensu and, one must add, as also often in Aristotle’s dialectical discussions» (pp. 23f.). I agree with the second sentence, but have reservations about the first. Runia 1992 and 1989 tried to show that that structure is present in chapters I, 20 (‘On the substance of the sun...’) and II, 25 (‘On the substance of the moon’) of Ps.-Plutach’s Epitomé, suggesting that a diaphônia is admitted between the position taking fire as the prevailing element in the sun and in the moon and the position taking earth as the prevailing element. The alternative suggestion however is that there five main positions involved, viz. those making recourse to one of the four (Empedoclean) elements (among them Xenophanes’ clouds involve air, Pythagoreans’ glass, being similar to ice, involves water), and the one by Aristotle that makes recourse to a fifth element. This has the advantage of not regarding the tenets by the Pythagoreans and by Aristotle as odd tenets. But on this interpretation there is no diaphônia, the presence of which is not suggested by even the slightest comment in those chapters. One can add, as a counterexample to Mansfeld’s and Runia’s thesis, that the chapter I, 24, which is about a topic, that of generation and corruption, which is particularly apt for propounding a diaphônia, viz. that between those who deny generation and corruption and those who admit them, in fact contains a third, intermediate position: that attributed to Empedocles and Epicurus, according to which they are processes that are to be reduced to the combination and separation of particles. There are cases where the diairesis involved looks like a tree, and this may or may not be combined with an opposition between two main positions53. There are, of course, examples that fit the schema propounded more neatly, but one does not get the impression they are prevalent. Anyhow, I do not believe I have exhausted this argument with such quick remarks, but perhaps they can help to make us aware that the issue of the methodology that is applied is quite complex and requires a very differentiated treatment.

  • 54 See however Mansfeld 1989 and 1992.

56The second issue involved concerns the supposition which one is naturally lead to make in this perspective (though it is not explicitly made by Mansfeld and his collaborators54) namely that, because the first full formulation of the method goes back to Aristotle, not only must its first full application in the field of doxography go back to his pupil Theophrastus, but all later applications must depend on this first application, thus tacitly excluding the possibility that it has been largely transformed in the course of time. Mansfeld himself admits that there were later developments of the method, as testified by certain works in the field of rhetoric and/or dialectics. Yet these developments are only a part of those that must have taken place, and which are unknown to us because the works containing them are lost. Thus the technique of diairesis certainly was of interest to the Stoics just as it was of interest to later Peripatetics and Academics. What excludes, then, that the application of this and other ‘dialectical’ techniques in works like Aëtius’ Placita was suggested at least in part by later handbooks expounding these techniques rather than by familiarity with their application in doxographical works to be attributed to Theophrastus? One difficulty with the suggestion that we should go back to Theophrastus is that none of his doxographical contributions known to us (especially the long passage given by Simplicius in his commentary of Aristotle’s Physics I, 2, and the De sensibus) looks very much like what we find in an Epitomé like that by Ps.-Plutarch, for they contain a relatively ample exposition of the contributions of each Presocratic under consideration e.g. on a comprehensive topic like the functioning of the senses (in general and singly), and not a succession of statements (often in just one or two sentences) by different Presocratics on a very specific topic (such as the functioning of the sense of sight). Another difficulty is illustrated in what follows.

57The third issue concerns the way in which this ‘dialectical’ method is applied in works like the Epitomé by Ps.-Plutarch, as only admitting that it is applied with sufficient consistency enables us to provide an analysis that permits us to draw inferences about the contents of its source. This assumption is particularly relevant in the case of that Epitomé, for it is the work that constitutes the basis for any attempt to reconstruct Aëtius, by admitting that what its author had done in most cases was to omit certain lemmata, while respecting the organization of the source for his information. Since part of the missing lemmata are to be found in the parallel passages that are present in Stobaeus and in other works (especially Theodoretus’ Graecarum affectionum curatio), a reconstruction of the Placita of Aëtius becomes possible that, for some chapters at least and for the general lay-out, is very close to the original.

58It is mainly to this third issue, but to some extent to the second or more generally to the issue of reconstructing Aëtius, that I shall now give my attention. One point of departure is the admission (by Mansfeld and his collaborators) that part of the ‘dialectical’ method that is applied in Aëtius’ Placita and in the Works that are based on it consists in grouping all the material (i.e. opinions or tenets by thinkers) around certain problems and question-types. These question-types go back to Aristotle and can be schematically formulated as follows: 1. Existence (‘whether something exists’), 2. Essence (‘what it is), 3. Quality (‘what sort of quality it has) or property in general, 4. Quantity (‘how many’, ‘how great’, etc.), 5. Place (‘where is it), 6. Cause (why this is so and so’). (I offer a somewhat changed formulation of the question-types as they are expounded by Mansfeld 1992, sect. VI.) That this part of the method or this technique (whether dialectical or not) has some sort of application to Ps.-Plutarch’s Epitomé cannot be denied, given the examples that can be given and that in fact have been given by Mansfeld and his collaborators. The questions however that can be raised are whether it remains essentially the same methodology as that propounded by Aristotle and whether it is applied in a reasonably consistent way.

  • 55 Cf. Lachenaud 1998, esp. pp. 42-4.

59If one looks at the picture given by Guy Lachenaud55 of the question-types introduced in the Epitomé, one notices that, while it remains much the same in the overall, it presents two important discrepancies with that given by Manfeld: the question of existence is completely left out, and the most frequent chapter-heading that is given (84 titles) is not a question at all, because it is of the neutral form «On X» (Περὶ X). Lachenaud rightly observes of this neutral formulation that «c’est celle qui se rapproche le plus du mode d’indexation des dictionnaires et il faut lire le chapitre pour se faire une idée de la question philosophique». But if the Aristotelian or Peripatetic method of dialectical questioning were still alive, at least in the sense of remaining in the background, its wide disappearance in the chapter-headings would be surprising. The other discrepancy is important too, for one has to stress that, on Aristotelian assumptions (assumptions that are evident in the first chapters of book II of the Posterior Analytics, which is the first source for the schema) question-type 1 comes before question-type 2, just as this second question-type comes before the others (except when question-type 6 coincides with question-type 2). If one keeps this in mind, one has the surprise that there is not even one chapter-heading in Ps.-Plutarch’s Epitomé that responds to question-type 1.

  • 56 For a detailed discussion of its contents see Runia 1996a.

60Mansfeld misleadingly suggests that ch. I, 7 responds to it, but in fact the title of the chapter is Τις θεός, which presumably responds to question-type 2, certainly not to question-type 1. The suggestion that existence is at issue comes from the content of the first part of this chapter (i.e. from its beginning to 881D, or §§ 1-10), which is clearly very different from its second part, which contains the standard doxography (‘Thales said that god is the intellect of the world’, etc.) that in fact responds to the title of the chapter, for it States what god is for each of the thinkers mentioned there. The first part of the chapter is a mixed bag, for it initially expounds (in some detail and with quotations) the positions of those who denied the existence of gods, then, assuming (without any argument) that there are gods, it is maintained (presumably from an Epicurean point of view, and rejecting alternatives) that they are too perfect to be involved in the things of the world, thus to have a providential function56. It does not contain a proper diaphônia about the existence of god (or gods), for one would expect that after the exposition of the positions of the atheists an exposition were given of the positions of those who maintain its existence.

61There are other anomalies along the same lines that can be detected in the Epitomé of Ps.-Plutarch. For instance, in book I there are six chapters in the case of which a first chapter with the title ‘On X’ is followed by a chapter with the title ‘On the essence of X’ (21: on time, 22: on the essence of time; 25: on necessity, 26; on the essence of necessity; 27: on destiny [ειμαρμένη], 28: on the essence of destiny). This way of proceeding would be rather satisfactory if the first chapter in this sequence consistently dealt with the question of the existence of X, while the second dealt with its essence, taking the existence for granted. Yet ch. 21 on time contains definitions of time which are hard to distinguish from those contained in ch. 22, and also ch. 27 on destiny includes definitions. Furthermore, the similar chapter 29 On chance (τύχη)’ is not accompanied by a chapter ‘On the essence of chance’ and clearly deals with the essence of chance, for it collects the definitions that have been given of it by Platon, Aristotle, Anaxagoras and the Stoics. To this group of chapters there correspond in Stobaeus single chapters with more extended titles (‘On divine necessity...’, ‘On destiny and the good order of events’, ‘On chance and spontaneity’, etc.). Not all of these titles are satisfactory (the necessity admitted by the atomists is not divine), but some of them are (the chapter on chance and spontaneity cointains a lemma about Aristotle’s position about the relationship between these two terms, and there is a corresponding passage in Ps.-Plutarch). It is not obvious that the chapter-headings in Ps.-Plutarch have to be preferred as more ‘original’.

62In some cases the most proper title for a chapter would be one that showed that it is about the difference between two terms, but this type of chapter (using the formula Τίνι διαϕέρει) appears only twice in Ps.-Plutarch (in I, 2 about the difference between principle and element, in IV, 12 about the difference between ϕαντασία and connected terms). It is not used in the case of the three following related terms: place (τόπος), void (κενόν) and space (χώρα), which are introduced in I, 18-20.

63There are other anomalies that can be pointed out, one of which is given by the contents of the chapter II, 1 ‘On the world’, for it puts together the question of its essence and denomination (Pythagoras and, partly, Empedocles), the question whether there is one world or more than one (Thales and his followers on one side, the atomists on the other), and the question whether the world and the universe (τò πᾶν) coincide (Diogenes of Apollonia and the Stoics). There is some coincidence in the contents between this chapter and I, 6 on the question ‘Whether the universe (or the whole: τό πᾶν) is one’, for some of the lemmata concern not only this question by itself but also the question of whether there is one world or more than one, and, in one case (that of Metrodorus of Chios, mentioned also in the other chapter, unless it is the Epicurean), only the second question is explicitly considered.

  • 57 The importance of this passage vvas stressed by Mansfeld 1990, 3141-3.

64Some of the topics now touched upon appear in a passage in Galen’s De locis affectis III, 5 which is of interest because it contains a list of issues that are explicitly said to be the subject of dialectical questioning57. I quote: «This also holds with coming into being and passing away, just as with the substance of the soul, and with the gods and with the providence and destiny, and with (the question) whether the world has come into being or has not come into being, and whether the universe is unlimited or limited, or whether there are many worlds or uncountably many or just one, viz. this one. » Mansfeld himself recognizes that there is a discrepancy between the ‘coalescence’ of topics in Epitomé II, 1 and their distinction in Galen, but does not recognize that the discrepancy must regard also the questions about the gods, about providence and about destiny, for these must be primarily about their existence. The passage confirms that even an ancient author would have recognized that the organization of the material in the Epitomé is unsound on various points. In any case, there is not one chapter which, even independently of the chapter-heading, explicitly formulates the question of existence about something and expounds contrasting opinions on that issue, as happens in Aristotle (e.g. in Physics II, 4 the question is raised as to whether chance exists, and contrasting opinions on the issue are expounded, in Physics III, 4 and IV, 6 and 10 something similar happens with the existence of the infinite, with that of the void and with that of time).

65More generally, one cannot rely wholly on the division of chapters, with their chapter-headings, and even on the division of books in Ps.-Plutarch, as it is not hard to show. At the beginning of book IV of his Epitomé there is a chapter on the flood of the Nile, which is introduced by a statement to the effect that, after a ‘perlustration’ of the parts of the world, one will enter into details. One expects, after that chapter, to which nothing corresponds in Stobaeus, an exposition of rather particular and peculiar phenomena, like that which we find in book VI in Lucretius (on the exhalations of the lake of Avernus, on founts, on the action of the magnetes...), but in fact the successive chapter is entitled “On the soul” and introduced a series of chapters on psychical phenomena. Another topic that should be connected with IV, 1, that of rivers, is missing too.

  • 58 Cf. Mansfeld 1990, 3189f.

66Another thing that one notices is that the first two chapters of book V, on divination and on dreams, have little to do with the ones that follow on the various aspects of animal reproduction: these could be treated as affections (πάθη) of the soul, but, surprisingly, the last and rather short chapter of book IV deals narrowly with the «bodily affections» (nothing corresponds to this in Stobaeus, for the ‘affections’ of ch. 60 are those of respiration). These two chapters should rather be connected with some chapters that are inserted at the end of book V, that is ch. 24 (‘How does dreaming take place, and how does death’), ch. 25 (‘Of which is dream and is death {an affection}, of the soul or of the body), ch. 30 (‘On health and on old age and on illness), and, in addition, with a chapter in another book (IV, 22 about respiration), for these all are topics which roughly correspond to the contents of Aristotle’ s Parva naturalia. But even this later group of chapters is interspersed with chapters of a different type, and, in particular, ch. 29 (‘On how fever occurs and whether it is an after-symptom) is quite exceptional and anomalous, since, because of its strictly medical nature, it is not appropriate in a work on Physical opinions, and has nothing corresponding in Stobaeus (where on the other hand there is a similar disorder in the chapter concerning the topics now mentioned or similar ones). Mansfeld58 is not unaware of the disorder in the last chapters of the Epitome, but tries to circumscribe it by treating this part as an ‘appendix’ to the work, not considering the obvious connection between some of these chapters and the first two chapters of book V.

  • 59 This is conceded by Mansfeld 1990, esp. 3084f.

67As to the title «On the soul» in book IV, this is too general, for it introduces a topic dealt with in a plurality of chapters («On the essence of the soul» would have been more appropriate). Furthermore, the contents of the chapter which it entitles are not separable from the contents of the next chapter (entitled «Whether the soul is a body and what is its essence»), since there is a disjunction between those who regard the soul as immaterial (ἀσώματον) and those who regard it as a body. That there is more justification in keeping this material together in one chapter, as it happens in Stobaeus, is rather clear59.

  • 60 In fact Diels 1878, in his edition of Aëtius, adopts the title ‘On the hegemonikon as Placita IV, 5 (...)
  • 61 The remaining part is about the identity of intellect (νοῦς) and soul (ψυχή) that is admitted by De (...)
  • 62 Cf. Mansfeld 1990, 3092, n. 138.
  • 63 Cf. Mansfeld 1990, 3117-21, with quotation of the Greek text.

68One misses, as rather naturally belonging to its context, the two distinct chapters ‘On the nature of men’ (I, 47) and ‘On the intellect’ (I, 48) that are to be found in Stobaeus, unless one supposes that the first represents implicitly (as it does explicitly in Theodoret) the title of the whole book, since it contains no doxography (but only quotations from Plato etc.). To the second can only partly correspond the chapter ‘On the hegemonikon’ in Ps.-Plutarch (as Diels assumed), for this concerns mainly the collocation of this part of the soul (the chapter-heading again is not quite appropriate) not its essence, and the ‘hegemonikon’ cannot be taken as coinciding with the intellect for all thinkers60. Mansfeld 1990, 3151, recognizes that part of the contents61 of the chapter ‘On the intellect’ in Stobaeus cannot be taken as regarding the ‘hegemonikon’ and that it has parallels in what is to be found in Theodoret V, 2862 and in Philo of Alexandria, De somnis I, 3163, but is reluctant to admit that it belonged, from the beginning, to a separate chapter concerning the intellect, though not denying that this is a possibility (cfr. the passage referred to in n. 61). He thinks that the question, formulated there, as to whether the intellect is penetrated from the outside (θύραθεν in Stobaeus and in Theodoretus, ἔξωθεν in Philo) is equivalent to the question ‘wherefrom does the soul come’ which is to be found in other sources (e.g. Cicero, Tusculanae I, 18: unde). But the former question goes back to Aristotle (cfr. De generatione animalium II, 3, 736b27-29) and concerne the intellect taken as distinct from the soul. (The same position is taken up with explicit reference to Aristotle by Nemesius, De natura hominisch. 1, 1.5 ff., espec. 1.16-17, as recognized by Mansfeld 1990, 3092, n. 138, without however admitting that it can only concern νοῦς; that the intellect is distinct from the soul for those who admit that it cornes from the outside is clear also in Stobaeus.)

  • 64 Cf. Mansfeld-Runia 1997, 186.

69Runia64 admits that this is the most likely explanation of what has happened, and also admits that, from the title adopted in Stobaeus for ch. I, 21, it is likely that a chapter ‘On where the world has its hegemonikon’ is missing in Ps.-Plutarch. He also admits that stray lemmata in Ps.-Plutarch and in Stobaeus make one expect a chapter ‘On the movement of the world’, not noticing that similar, though not stray, lemmata also make one expect a chapter entitled ‘Wherefrom the stars are nourished’ which has its parallel in a chapter entitled ‘Wherefrom the world is nourished’ (II, 5 in Ps.-Plutarch’s Epitomé, part of the topics indicated in the title of ch. 21 in Stobaeus). These lemmata in Ps.-Plutarch’s Epitomé belong to chapter II, 17 entitled ‘Wherefrom the stars get their light’ (while in Stobaeus they belong to ch. 24 ‘On the substance of the stars etc.’), and are put by Diels 1878, 346, under the same title in his edition of Aëtius, but it is sufficiently clear that this title is not appropriate.

70Going back to book I, one would expect the chapter 16 ‘On the division of hodies’ be kept close to those ‘On bodies’ (12) and ‘On minimal parts (ελάχιστα)’ (13), when in fact it is kept separate from them by chapters ‘On figures’ and ‘On colours’. In Stobaeus there is instead just one chapter ‘On bodies and on their division and on the minimal part’, which is not satisfactory since what is said of bodies is independent of what is said of their division, but which follows a logical sequence which is missing in Ps.-Plutarch.

  • 65 Cf., for a recent contribution on this point, Runia 1996b.
  • 66 A partial recent recognition of this fact is to be found in Runia 1999, 196f., where he says that « (...)

71Another important anomaly or complication is the following: it is usually admitted that in Stobaeus there are either whole chapters or parts of chapters which come from a source that is different from Aëtius and which, following a suggestion by Diels, is supposed to be Arius Didymus65. What is not usually recognized is that there are chapters or parts of chapters in the Epitome of Ps.-Plutarch which also must come from a source, or from sources, which are different from Aëtius66. One of these pieces is the part of chapter I, 7 discussed above concerning gods. The chapter immediately preceding it, entitled «Wherefrom men have their notion of the gods», appears to contain Stoic doctrine on that issue (it is included in Stoicorum veterum fragmenta II as fr. 1009). The atomistic description of the formation of the world, in I, 4, under the title «How has the world been constituted», must be of Epicurean origin (it is included by Usener as fr. 308 in his Epicurea). Chapters 1 and 2 (entitled: ‘what is nature’ and ‘In what principle and element differ (from each other)’) are more introductory than doxographical, while chapter 3 (entitled: ‘On the principles, what they are), though being typically doxographical, contains a much wider exposition (occasionally accompanied by quotations) of the positions that are considered than that given elsewhere. Further, in book III chapters 5 and 18 contain (anonymous) Peripatetic meteorological material, and in book IV chapters 11-12 and 21 contain (explicit) Stoic epistemological and psychological material. If one considers that the first book includes a preface on Stoic (cfr. Stoicorum veterum fragmenta II, 35) and Peripatetic conceptions of philosophy and that, in its general title, omits the word epitomé which, with the exception of book IV, is included in the title of the single books, one has to wonder if the author has not changed his intentions at some stage in preparing it and, giving up the original plan of writing a work of wider scope, based on use of various sources, has resorted to excerpting from a work like Aëtius’ Placita.

  • 67 Cf. Mansfeld-Runia 1997, 184-7, where they tend to take for granted the correctness of Diels’ hypot (...)

72In conclusion, the fact that there are anomalies is not too surprising, but the question of course is how many anomalies can be allowed before recognizing that it is not just a matter of details but there is some serious defect in the organization of the material that excludes a very satisfactory reconstruction of the Aëtius’ Placita, if this work is the main source on which Ps.-Plutarch draws. Surely more anomalies would emerge on a less selective examination than that conduced by me above. Another, connected, conclusion, is that the organization of the chapters in Ps.-Plutarch is not constantly, or almost constantly, to be preferred to that present in Stobaeus, when it is rather clear that, if that organization was methodically rather sound in Aëtius, both deviated from it in various ways, while, if it was messy too, we cannot tell where this mess is more faithfully reflected. Mansfeld and Runia share the former conviction with Diels67, but none of them offers an adequate justification of it. Of course it is not to be denied that an approximate reconstruction of Aëtius’ Placita is feasible, for the concordances between Ps.-Plutarch’s Epitome, Stobaeus’ collection and (to some extent) Theodoret’s work are too many not to point to a common source. But if in this, which is by far the most favourable case for the reconstruction of the doxographical tradition, we encounter so many limitations, what can be said when it is a matter of speculating, using very partial and scattered data, about the organization of the Vetusta placita and other similar works that can be postulated for the period between Theophrastus and Aëtius?


Alfieri, V.E., 1953, Atomos Idea. L’origine del concetto dell’atomo nel pensiero greco, Florence.

Baffioni, C., 1981, «Salomon Jakovlevic Lur’e interprete di Democrito», Elenchos 2, 1981, 161-92.

Bailey, C., 1928, The Greek atomists and Epicurus. A Study, Oxford.

Barnes, J., 1987, ed., Early Greek Philosophy, Harmondsworth.

Bollack, J., 1969, ed., Empédocle, 2. Les origines. Edition critique, Paris.

Burkert, W, 1998, «Pythagoreische Retraktationen: Von den Grenzen einer möglichen Edition», in Burkert et alii, 303-19.

Burkert et alii, 1998, Fragmentsammlungen philosophischer Texte der Antike, Hrsg. Burkert, Gemelli, Matelli, Orelli, «Aporemata» 3, Göttingen.

Burkert, W., 1999, «Diels Vorsokratiker. Rückschau und Ausblick», in Hermann Diels (1848-1922) et la science de l’antiquité. Entretiens préparés et présidés par William M. Calder III and Jaap Mansfeld, Vandoevres, 169-97.

Decleva Caizzi, F., 1980, «Democrito in Sesto Empirico», in Democrito e Tatomismo antico, cur. F. Romano, Catania, 393-410.

Diels, H., 1879, Doxographi graeci. Collegit recensuit prolegomenis indicibusque instruxit H.D., Berlin.

Fortenbaugh, W.W., and Steinmetz, P., 1989, edd., Cicero’s Knowledge of the Peripatos, «Rutgers University Studies in Classical Humanities» 4, New Brunswick N.J.

Fortenbaugh, W.W., and Gutas, D., 1992, Theophrastus. His Psychological, Doxographical and Scientific Writings, «Rutgers University Studies in Classical Humanities» 5, New Brunswick.

Fortenbaugii, W.W., Huby, P.M., Sharples, R.W., and Gutas. D., 1992, edd., Theophrastus of Eresos. Sources for his Life, Writings, Thought and Influence, Part One: Life, Writings, Various reports, Logic, Physics, Metaphysics, Theology, Mathematics, Leiden (abbrev. FHSG).

Gemelli Marciano, L., 1998, «Wörtliche Zitate aus Demokrit in der skeptischen Ueberlieferung», in Burkert et alii, 106-128.

Gottschalk, H., 1967, Review of Steinmetz 1964, Gnomon 39, 17-26.

Hölscher, U., 1969, Parmenides, Vom Wesen des Seienden. Die Fragmente, Frankfurt/Main.

Huffman, C. A, 1993, Philolaus of Croton. A commentary on the fragments and testimonia with interpretative essays, Cambridge.

Kahn, C., 1979, The art and thought of Heraclitus. An edition of the fragments with translation and commentary, Cambridge.

Kühn, C.G., 1828, ed., In Hippocratem de natura hominis commentarius, in Claudii Galeni Opera Omnia, Tomus XV, Leipzig.

Lachenaud, G., 1998, «Le De placitis philosophorum manifeste-t-il une visée encyclopédique?», in Burkert et alii, 41-61.

Laks, A., 1997, «Du témoignage comme fragment», in Collecting fragments / Fragmente sammeln, edited by Glenn W. Most, «Aporemata» 1, Göttingen, 237-72.

Lee, H.D.P., 1936, ed., Zeno of Elea, A text. with Translation and Notes, Cambridge.

Luria, S., 1970, Democritea: collegit, emendavit, interpretatus est Salomo Luria, Leningrad.

Mansfeld, J., 1989, «Gibt es Spuren von Theophrasts Phys. Op. bei Cicero?», in Fortenbaugh and Steinmetz, 133-58,

Mansfeld, J., 1990, «Doxography and dialectic: The Sitz im Leben of the Placita », Aufstieg und Niedergang der Römischen Welt II.36.4, 3056-229.

Mansfeld, J., 1992, «Physikai doxai and problemata physika from Aristotle to Aëtius (and beyond)», in Fortenbaugh-Gutas, 63-111 (a shortened version in French translation in Revue de Métaphysique et de morale, 1992, No. 3 on «La doxographie antique »).

Mansfeld, J., 1998, «Doxographical Studies. Quellenforschung, Tabular Presentation and other varieties of comparativism», in Burkert et alii, 16-40.

Mansfeld, J., Runia, D.T., 1997, Aetiana: The Method and Intellectuel Context of a Doxographer, I: The Sources, «Philosophie Antiqua» 72, Leiden.

Marcovich, M., 1975, Review of Luria 1970, in Archiv für Geschichte der Philosophie 57, 60-3.

Marcovich, M., 1978, ed., Eraclito: Frammenti, «Biblioteca di studi superiori» 54, Florence.

Morel, P.-M., 1996, Démocrite et la recherche des causes, Paris.

O’Brien, D., 1982, «La taille et la forme des atomes dans les systèmes de Democrite et d’Epicure», Revue Philosophique de la France et de l’étranger 172, 187-203.

Osborne, C., 1987, Rethinking Early Greek Philosophy. Hippolytus of Rome and the Presocratics, Ithaca.

Runia, D.T., 1989, «Xenophanes on the moon: a doxographicum in Aëtius», Phronesis 34, 245-69.

Runia, D.T., 1992, «Xenophanes or Theophrastus? An Aëtian doxographicum on the Sun», in Fortenbaugh-Gutas, 112-40.

Runia, D.T., 1996a, «Atheists in Aëtius. Text, Translation and Comments on De Placitis 1.7. 10», Mnemosyne 49, 542-76.

Runia, D.T., 1996b, «Additional Fragments of Arius Didymus on physics», in K.A. Algra, P.W. van der Horst and D.T. Runia, eds., Polyhistor: Studies Presented to Jaap Mansfeld, Leiden, 363-81.

Runia, D.T., 1999, «The Placita ascribed to Doctors in Aëtius’ Doxography on Physics», in Ancient Histories of Medicine: Essays in medical doxography and historiography in classical antiquity, ed. by Ph.JU. van der Eijk, «Studies in Ancient Medicine» 20, Leiden, 189-250.

Rütten, Th., 1992, Demokrit lachender Philosoph und sanguinischer Melancholiker, «Mnemosyne» Supplement 118, Leiden.

Salem, J., 1996, Démocrite. Grains de poussière dans un rayon de soleil, Paris.

Sharples, R.W., 1998, Theophrastus of Eresos. Sources for his Life, Writings, Thought and Influence. Commentary volume 3.1: Sources on physics, by R.W.S., with contributions by D. Gutas, Leiden.

Smith, M.F., 1993, Diogenes Oenoandensis: The Epicurean Inscription, ed. with intr., transl. and notes, «La scuola di Epicuro» suppl. 1, Naples.

Smith, W.D., 1990, Hippocrates: Pseudoepigraphic writings. Letters, Embassy, Speech from the Altar, Decree. Edited and translated with an introduction, Leiden.

Steinmetz, P., 1964, Die Physik des Theophrast, «Palingenesia» 1, Bad Homburg.

Taylor, C., 1999, The Atomists: Leucippus and Democritus. Fragments: A Text and Translation with a Commentary, «The Phoenix Presocratics» 5, Toronto.

Zhmud, L., 1997, Wissenschaft, Philosophie und Religion im altem Pythagoreismus, Berlin.

Zhmud, L., 2001, «Revising doxography: Hermann Diels and his critics», Philologus, 145/2, 219-43.


1 I am grateful for observations made to one of the drafts of this paper to C. Huffmann, L. Gemelli Marciano, A. Laks, J. Mansfeld, D. Sider and L. Zhmud.

2 In some cases, as here and in the title, I talk simply of Democritus, since I do not think that Leucippus’ contribution to the history of thought can be distinguished from that of Democritus, though I do not share some scholars’ doubts about his very existence.

3 Meanwhile I had news of an edition of Democritus to be prepared by Laura Gemelli Marciano under the supervision of Walter Burkert but this would be limited to his physics and is not yet far advanced.

4 The provisory edition I made available to the Colloque in Lille is the result of such an act of will, but it is in need of some revision and should be published after the volume of translations (probably by the Edizioni di Storia e Letteratura, Rome). As to this other volume, which in various ways is complementary to it, a publication for the year 2002 is foreseen in a collection of the Academy «La colombaria» published by Olschki of Florence.

5 A few further observations below (these do not concern his commentary). For a general criticism of Luria’s edition, with comparison with that of Diels, see Marcovich 1975. Some interesting observations are to be found in Baffioni 1981.

6 Not all of the changes that Kranz introduces in the texts, in their translation and in their apparatus are, in my view, an improvement on what Diels had done, but discussing and illustrating this point would take long. Some differences are pointed out in the notes to my translation or in the apparatus to the texts.

7 Cf. the edition, with the abbreviated title Aetii plac., given by Diels 1879, pp. 273-444.

8 On these points see Burkert 1999, esp. 178-80.

9 That this is Aristotles interpretation of their position is very clear in De caelo III, 1, 298b21 ff., the suggestion being that Parmenides and Melissus could not conceive of anything but sensibile substance-a form of sensism is attributed to the former in Metaph. IV, 5, 1009b21-5-but hypostatized it because they realized that knowledge has the immutable as its object, and, in addition, because they excluded change as being for them a passage from not-being to being. Their position is presented as a consequence of their failure to account for change, thus as a sort of misguided evasion from physics, also in Phys. I, 8, 191a23-31 and in Metaph. I, 3, 984a29 ff.

10 Text and translation can be found in Osborne, 1987, 328-37. It has to be admitted that Heraclitus is presented much more as the typical physicist and cosmologist in the account of Diogenes Laertius, IX, 7-11 (= 22 A 1 DK), which passage is taken by Diels (in his Vorsokratiker) as going back to Theophrastus’ Opinions of the physicists (but not quoted as a fragment from it in his edition of the Physicorum opinion es included in Doxographi graeci).

11 For more on the problems raised by an edition of the Pythagoreans see Burkert 1996.

12 This de facto applies to the second part of the poem, e.g. by using test. 37 (Aetius II, 7, 1) to fill the gap between frs. 8-9-10-11 and frs. 12-13.

13 Cf. Taylor 1999. The selection of the testimonies in this collection is rather restrictive (some of them are only referred to, but Taylor rarely goes beyond what is to be found in Luria), and this in itself favours Übersichtlichkeit, but I do not have the impression that the order adopted is always very convincing, even given that the author (as he states in the préfacé) «ordered them as seemed to him best to exhibit the structure of the atomists’ doctrines». He does not do the same for the ethical fragments, since the order followed (as in Diels’ Vorsokratiker) in quoting them is that of Stobaeus, and this probably has little to do with Democritus’ own intentions; furthermore, this group is kept separated from those belonging to the «Democrates collection» that are collocated in an appendix. Of course having everything at hand, including a commentary that clarifies various points, in the limits of one volume remains very convenient.

14 Indeed there is only one sure reference (Taurus apud Philoponus, De aeternitate mundi contra Proclum VI, 8, 145.20-24 Rabe = fr. 11 of Diels’ edition of Theophrastus’o piniones = 241A FHSG), if we leave out the two references (if such they are) to its Epitomé and we reject Diels’ attempt to recognize the same work in the Περὶ (τῶν) φυσικῶν (see below and also see the discussion of the evidence in Steinmetz 1964, 334-51).

15 From now on I add the abbreviation «Le. » (to be kept distinct from «Lu. » for «Luria») when giving the number of a passage in my collection, and, whenever available, I permit the number in Diels-Kranz’s collection.

16 On the parallels between Aristotle’s treatment of the earth in De caelo II, 13 and Ps.-Plutarch’s Epitomé, III, 9-15, see Mansfeld 1992, sect. VII.

17 A reference to the monograph in two books on Anaxagoras is made by Simplicius in his comm. on Phys. I, 4, p. 166.15 ff. (= 59 A 41 DK), and this work must have been his main source for the quotations he gives. Presumably the title he adopts corresponds to two titles of works on Anaxagoras in one book given in Theophrastus’ catalogue (cfr. Diog. Laertius V, 42).

18 The title Φυσικά (excluding φυσικοί as a possible reading of the genitive, see below) is used by Alexander of Aphrodisias in his comm. to the Metaph., 31.7f. Hayduck, concerning Parmenides’ position (= fr, 6 of Diels’ed. of the Physicorum opiniones = 28 A 7 DK), by Diogenes Laertius IX, 22, equally concerning Parmenides’ position (= fr. 6a = 28 A 1 DK), by Simplicius in his comm. to the De caelo, 564.24 Heiberg, concerning Democritus’ position (= fr. 13 = 68 A 120 DK = 47.1 Le), by Aëtius II, 20.3 again on Xenophanes (= fr. 16 = 21 A 40 DK, but this is more doubtful, because based on a conjecture by Diels).

19 As Diels assumed, taking it, in a questionable way, as equivalent to the Περὶ φυσικών ἐπιτομή in 2 books mentioned in the catalogue of Theophrastus’ works (cfr. D.L. V,46) (see Diels 1878, 103).

20 Cf. fr. 5a Diels (= 231 FHSG).

21 This is how Mansfeld and Runia, 1997, 183, n. 203 (with reference to other contributions by Mansfeld), understand the passage.

22 Cf. his commentary to Aristotle’s Physics, 115.11f. Diels, concerning Parmenides (= fr. 7 = 234 FHSG = 28 A 28 DK), and 154.14-7 concerning Anaxagoras in relationship to Anaximander (= note to fr. 4 = 228B FHSG = 12 A 9a DK). The shorter ‘Ιστορία, surely for the same work, recurs in 149.32 concerning Anaximenes (= note to fr. 2 = 226B FHSG = 13 A 5 DK).

23 This analogy was already propounded by Diels 1879, 102.

24 This is normally quoted by him as Φυσικά e.g. at the beginning of his commentary to Aristotle’s Physics (cfr. 20.19f. = 143 FHSG, and cfr. 9.7 = 144B). A distinction between a Naturalis historia and a Naturalis auditus (= Φυσική άκρόασις) was apparently recognized also by Priscian in his Solutiones (cfr. Suppl. Arist. I, 2, 43.3), as already stressed by Diels 1878, p. 102 (following Usener).

25 This is the position that was already suggested by Gottschalk 1967, is contemplated by Sharples 1998, 12, and is most recently defended by Zhmud 2001.

26 Hence Kühn 1828, 25, translates as follows: «... in quibus naturales opiniones in compendium redegit».

27 Cf. Mansfeld 1992, 64f., and n. 10. Also Sharples 1998, 10, points out that it would have been easy, if the title had the other meaning, to make this clear e.g. by using περί with the accusative.

28 This has been pointed out by Zhmud 2001, who adduces evidence. (In what follows I stick to the convention of using the title Physicorum opiniones).

29 Mansfeld 1996, 23f. (also cfr. Mansfeld-Runia 1997, xix-xx).

30 Cf. Stobaeus I, 14, 1 f(2-3) = 68 A 47 = 64.7 Le. ; Ps.-Plutarch, Placita I, 3, 9 = 68 A 47 = 9.1 Le.

31 E.g. that he introduced the clinamen as an uncaused deviation of the atoms from their course to avoid determinism, cf. Cicero, De finibus I, 6, 17-20 = 9.3 Le. [this is one case where the quotation in Diels and Kranz is inadequate, being limited to § 17, cfr. 68 A 56], De fato 10, 22 and 23 = 281 Us. = 76.2 Le., and De natura deorum I, 25, 69 = 281 Us. = 76.3 Le. [both omitted in Diels and Kranz].

32 Not made by Democritus. (It is likely that Epicurus himself is rejecting the determinism he attributes to Democritus, without making his name, in a fr. [= 68 A 69 = 74.2 Le.]) of his Περὶ φύσεως).

33 On this example cf. O’Brien 1982.

34 This only happens in some of his letters of which fragments have remained (2 based on Philodemus, cfr. 104 and 113 Arrighetti = 0.8.34 and 0.8.35 Le). The fact that in one of these passages he seems to mention works by Democritus conferms what is likely anyway, i.e. that he was acquainted with some of them at least, while this probably has to be excluded for later Epicureans.

35 Cf. Epistula ad Pythoclem §§ 89-90 (= 67 A 24 DK = 81.2 and 80.3 Le.) on cosmology; Epistula ad Herodotum, § 49 (= 117.1.1, om. DK), on the mechanism of sight; ibid., § 43 (= 90.3.1, om. DK), on the nature of sound; ibid., §§ 42-43 (= 40.2, om. DK), and §§ 55-56 (= 68 A 43 = 40.3), on the issue touched upon above of the variety of figures; further, the fragm. from the Περὶ φύσεως mentioned above (n. 32).

36 On the composition of the soul, cf. III, 370-3 (= 68 A 108 DK = 103.6 Le.)

37 Cf. fr. 6, coll. II 9-14, III 1 (= 5.12 Le., om. DK), fr. 7, col. II 2-14, III 1-2 (= 8.4, om. DK) and fr. 54, col. II (= 68 A 50 = 76.1), fr. 10, coll. III 14 IV V (= 110.6), fr. 43, coll. I 12-14 II (= 110.7), e fr. 9, col. VI 3-14 (= 110.8). The last three are the new fragments on the nature of dreams. (I adopt the numeration of Smith’s edition.)

38 Perhaps it is not a chance that Diogenes Laertius’ work concludes with Epicurean doctrine and that some pieces in Ps.-Plutarch’s Placita (like the anonymous cosmogony in I, 4; other examples below) must be Epicurean. Further, a contribution to doxography can be found in Lucretius, book I, vv. 635-920.

39 I differ from Diels-whose quotation (68A112) starts at 1009b7-and from other scholars in admitting that the whole passage from 1009a38 concerns Democritus, but I cannot argue this point here.

40 That Arcesilaus used to appeal to the authority of other philosophers-though of these only Socrates coincides in both lists-to legitimate his position is confirmed by Plutarch in Adv. Col 26, 1121E-1122A.

41 Because the quotations he gives are rather short, when they are given at all, and in Adv. math. VII, 140 (= 68 A 111 and 59 B 21a DK = 60.1) he attributed to Democritus, evidently misled by his source, something that belongs to Epicurus (for details I must refer to the notes to my translation).

42 These are the two passages from De medica experientia: IX,5 [p. 99 Walzer = Nachtrag in vol. II DK, p. 423 = 60.2 Le., and XV, 7-8 = 68 B 125 = 60.3 Le.], the second of which contains two quotations of Democritus.

43 Because certain misunderstandings of his on Democritus’ position show lack of information, but this is a point I cannot illustrate here.

44 This conclusion, not extended to Galen and to Plutarch, was already drawn by Decleva Caizzi 1980, and it is extended to Galen but not to Plutarch-except as a source of information on Colotes-by Gemelli Marciano 1998.

45 I cannot believe that Democritus wrote a book on Pythagoras, of which otherwise we know nothing, and when no conspicuous Pythagorean influence is detectable in his thought, unless the supposition is made that this work concerned Pythagoreanism in a critical way. (For a different opinion see Zhmud 1997, 39-40 and passim.)

46 Cf. Nat anim. VI, 60 (= 68A150a DK = 96.8 Le.), XII, 16 (= 68A151 = 96.7 Le.), XII, 17 (= 68A152 = 97.3 Le.), XII, 18, 19, 20 (= 68A153, A154, A 155 = 98.5, 98, 6, 98.7 Le.), IX, 64 (= 68A155a = 99.3 Le.)

47 Some of the sources of Aelian have been studied by M. Wellmann in articles that appeared in Hennes 1891, 1892, 1895, 1896, and by R. Keydel in an article that appeared in the same review, 1937, but, unless something has escaped my attention, nothing useful for determining the source of the Democritean passages is contained in them.

48 Cf. Runia 1999.

49 One can compare Stob. IV, 6, 19 (= 68B267 = 156.1 Le.), III, 1, 45 (= 68B47 = 156.3 Le.), II, 9, 3 (= 68B174 = 157.6 Le.), IV, 7, 13 (= 68B268 = 166.4 Le.), II, 1, 12 (= 68B169 = 170.1).

50 Luria omits one of them, except for a reference to it, while Diels omits them all, except for a quotation of a little bit from one of them (cf. 68 A 55 DK).

51 I can only give one example (other examples in the introd. to my collection): among the passages which treat chance as an occult cause, that is 73.4, 73.4.1, 73.4.2, 73.4.3 Le. [= Aëtius I, 29.6-7 + Theodoret VI, 15 in apparatus, Dox. p. 326], only that by Theodoret mentions Democritus, and this is the variant that is quoted by Diels under Aëtius’ name as 59A66, referred to in 68A70.

52 One cannot take for granted that he agrees with Plato in this. In fact he discusses this procedure not in his handbook of dialectics; the Topics, but in the Posterior analytics and De partibus animalium, book I.

53 One case is that discussed by Mansfeld himself, 1990 (p. 3097), of the location of the ruling part (hegemonikon) of the soul: each of the main opposition positions (in the head, in the chest) admit an internal articulation which is tree-like; other cases are discussed by Laks 1997, who rightly points out that the topics which the opinions are about make a difference («le fait est que toute matière-la matière étant ici les opinions tenues sur un sujet determine – n'est pas «dialectisable» au même degré», p. 261).

54 See however Mansfeld 1989 and 1992.

55 Cf. Lachenaud 1998, esp. pp. 42-4.

56 For a detailed discussion of its contents see Runia 1996a.

57 The importance of this passage vvas stressed by Mansfeld 1990, 3141-3.

58 Cf. Mansfeld 1990, 3189f.

59 This is conceded by Mansfeld 1990, esp. 3084f.

60 In fact Diels 1878, in his edition of Aëtius, adopts the title ‘On the hegemonikon as Placita IV, 5, but in this chapter no lemma in Ps.-Plutarch corresponds to a lemma in Stobaeus (pp. 391f.).

61 The remaining part is about the identity of intellect (νοῦς) and soul (ψυχή) that is admitted by Democritus and others.

62 Cf. Mansfeld 1990, 3092, n. 138.

63 Cf. Mansfeld 1990, 3117-21, with quotation of the Greek text.

64 Cf. Mansfeld-Runia 1997, 186.

65 Cf., for a recent contribution on this point, Runia 1996b.

66 A partial recent recognition of this fact is to be found in Runia 1999, 196f., where he says that «it would be a mistake to regard the work as unified in approach and style. The first seven chapters of Book 1, for example, contain a number of discursive sections which deviate from the rest of the work, and there are also some unusually long descriptive passages in Book 3 and 4».

67 Cf. Mansfeld-Runia 1997, 184-7, where they tend to take for granted the correctness of Diels’ hypothesis, and just suggest some modifications on specific points.


© Presses universitaires du Septentrion, 2002

Conditions d’utilisation :