We are grateful to Professor Richard Janko for his speedy response to the publication of our edition of the Derveni papyrus and for his many kind remarks and observations.
J. opens his review with the impressive statement that “the Derveni papyrus is the most important text relating to early Greek literature, science, religion and philosophy to have come to light since the Renaissance”. This is a favorite obiter dictum of his, and he gives voice to it often (e.g. in the first paragraph of his review of Betegh’s edition of the papyrus, BMCR 2005.01.27), but surely it is hyperbolic. The Derveni papyrus is an important text, but let us not forget the old Greek adage μηδὲν ἄγαν.
We would like to clear up first a few misunderstandings that may arise from some of J.’s comments about the papyrus itself. His experience with the Herculaneum papyri is obviously undisputed, but it has led him astray when commenting on our papyrus. He is probably right in his hypothesis that the core of the roll was recovered first, but he is wrong when he opines that “the sequence in which Fackelmann removed the pieces had been jumbled or lost”, unless he means that Fackelmann himself jumbled the sequence — which he certainly did, even when dealing with the innermost part of the cylinder, i.e. the section which suffered the least damage. All 266 fragments were separated during Fackelmann’s one visit to the Archaeological Museum of Thessaloniki, and all were mounted between glass by him. It immediately became obvious to Kapsomenos when he first saw the papyrus, as it did to Tsantsanoglou and Parássoglou when they worked on it later, that Fackelmann, who did not leave any detailed notes on the sequence of removal, had not placed the majority of the pieces consecutively in the order in which he had separated them, that he had sometimes joined with adhesive tape pieces that certainly do not belong together, and that he had placed a largish fragment (fortunately a piece of the unwritten last section of the roll) upside down. All these facts have caused, and will continue to cause, serious difficulties. After Fackelmann’s otherwise brilliant intervention, no one has interfered with his mounting of all fragments in nine glass frames. In our publication of the only report Fackelmann submitted to the Museum authorities (pp. 4-5) we indicate, with insertions in brackets, the frames containing the layers that came from each portion of the roll as we identified them. We have not succeeded in determining the exact order in which Fackelmann separated all layers and arranged them on glass; the smaller fragments, coming from all parts of the roll, were certainly not deposited in any kind of order. For each frame, therefore, we have numbered the fragments consecutively, from left to right and from top to bottom, so that we, and future editors, might be able to refer to them.
It is true that we “do not openly acknowledge the effect of flattening carbonized papyri”; nor do we acknowledge it in any other way as regards our fragments. The Derveni papyrus has not behaved at all as the carbonized papyri from Herculaneum did. There has been no crushing of fibers or any rapid deterioration, as J. suggests; and there has certainly been no fading of the ink. Every time something needed to be checked, exactly the same amount of text could be read as the time before. The overall condition and appearance of the papyrus have not changed over the years. However, a minimal fraying at the edges has occurred in a few fragments, and this is where old photographs are of great importance. Believing that the Derveni papyrus must have behaved like the Herculaneum rolls, J. has misinterpreted some of our statements, and we are perhaps to blame for not being explicit or clear enough. For example, when we write, with reference to fr. G 1, that more letters were visible when Makaronas photographed the two largest sections of the roll before the separation of their layers (p. 70), we do not mean that the ink has faded or that the surface has deteriorated; we only mean that, when that particular layer was detached, its edges suffered some damage, as a comparison of the same fragment in plates 1 and 5 makes clear, especially at lines 5 and 11. Any differences seen in various photographs of a detached fragment are due to the skill of the photographer, the technique employed, and the conditions under which the photograph was taken. There has been no “continuing deterioration of the papyrus” that “went unrecognized”.
Naturally, we are very gratified when an expert like J. remarks that “one must […] admire the skill” with which we managed to combine 153 fragments to form a text of 26 columns. He is astonished, however, to learn that 42% of the total number of fragments have been left unplaced. Has there ever been in the history of papyrology such a method of calculation, by which the large and the tiny, the significant and the unimportant, the written and the blank are all considered equal and counted together? This reminds one of the farmer whose only possessions were a bull and a frog, and who, upon the demise of the latter, exclaimed in agony that he had lost half his livestock. Had J. wished to give a relevant, instead of an impressive but meaningless, percentage, he should have calculated the percentage of the comparatively minute amount of text that we did not manage to place; he would then have to use a one-digit number, and a small one at that. His observation too that “some [of the unplaced fragments] are as wide as half a column” is downright erroneous and unfair; there is no such fragment. He should also have borne in mind how difficult, if not impossible, it is to place a fragment above or below another piece without the control of the vertical fibers of their outer side. Anyone, however, who works on the Derveni papyrus has to labor without recourse to such a control; all existing photographs are of the written side only, since they were taken after the fragments had been placed on a thick, white paper, and then mounted.
Commenting on our Introduction, J. complains that he finds no report of his “central contention that most of the surviving treatise is a digression, meant to prove that taking traditional rites and texts literally is an impediment to true faith. So far as I know, this is the only hypothesis thus far advanced that explains the point made in cols. 5 and 20”. There is a host of attempts at explaining this point so as to shed light on how the Derveni author conceives of the relationship between such unlikely bedfellows as cult, religion, and cosmology. None, however, of these hypotheses is explanatorily privileged over the rest, and the markedly different hypothesis put forth in our Introduction, VI.9-16, is skeptically designed to show just that. At any rate, we find it hard to believe that such a large hunk of text is only a digression, even if we posit that the treatise was thrice the size of its surviving portion.
J. remarks that Kouremenos “rightly assigns [the Derveni author] to the school of Anaxagoras”. Although not disputing the obvious fact, granted by all commentators, that the author knew and was influenced by the thought of Anaxagoras, Kouremenos does not assign him to Anaxagoras’ or anyone else’s particular school. What he does is to compare and contrast the Derveni author’s cosmology with the various post-Parmenidean cosmologies, Anaxagoras’ included. Given the similarities among the members of this numerous family, it would be rash to affirm that the eclectic Derveni author belongs to a particular school.
J. misconstrues the skepticism expressed in the book about the possibility of identifying the Derveni author with any degree of certainty. What is written on p. 59 is that “since the author could be identified only on the basis of evidence from the work, deciding between incompatible readings of a fragmentary work in this way would turn into reinforcing one’s own preferred reading by attaching to the work a suitable name. It is difficult to see how this vicious circle could be broken.” J. retorts that “the editing of the text will in the first instance be based on its language and on the well-established principles of textual criticism, not on hypotheses of authorship.” Quite so. The context, however, shows that at this point Kouremenos is not writing about the editing of the Greek text but about its understanding, and that by “readings” he means “interpretations”.
Coming to the text, we are sorry that J. was confused by our editorial conventions, and that we did not always conform to his ideal use of the sublinear dot. There is however no need to resort to an “intensive study of the literature on the papyrus” to understand our practice. A glance at the papyrological apparatus and at the available images will enable anyone to add or to remove as many sublinear dots as she or he thinks proper. J., however, is right in correcting an oversight of ours; in our description of the first letter in col. 5,9, read “practically nothing” for “nothing”, and let the dot stand.
Far more serious is our “simple but bizarre expedient” not to include an apparatus criticus. To the publication of his “interim text” (ZPE 141, 2002, 1-62) J. has appended an invaluable apparatus, which we could have easily corrected, brought up to date and appended to our text. But if such an apparatus was necessary for the establishment of his text and perhaps still retains some historical interest, what need is there for it now, when the reader has plates of all fragments and a detailed papyrological apparatus? What would be the point of listing our former, and now rejected, published readings and supplements? Or of cataloguing the readings and supplements suggested by scholars “working blind”, to use J.’s words, without photographs and with no knowledge of the length of the gaps? Just to show where blindness or disbelief can lead? When we owe the reading of even one letter to a particular scholar, we own the fact (p. 69, n. on line 8); otherwise all readings are due to Kapsomenos and/or ourselves. J. states that by not appending an apparatus criticus we have “contrived not to acknowledge that scholars other than [our]selves have toiled to reconstruct this text.” This is certainly not an accurate description of our intentions. The commentary in the second part of our book acknowledges all foreign supplements adopted by us (e.g. M. L. West’s in the Orphic verses), unless we inadvertently missed one (for which we are truly sorry), and discusses or mentions all those we considered worthy of discussion or mention. Should we have listed and taken stock of all proposed readings and supplements, including the impossible, the improbable, the extravagant, and the ungrammatical ones, just in order to portray the extensive toil of fellow scholars? J.’s additional remark that our “openness to supplements other than [our] own ceased at least a decade ago” is perhaps to be explained by his dissatisfaction with the fact that “the textual proposals of the present writer are either rejected or (mostly) ignored”. We took all his contributions of the last decade into consideration; we reported those we thought worth reporting (albeit we admit that we did not adopt any), while we tacitly ignored all those that we thought unworthy of mention, mistaken, or unsupported by the physical evidence of the papyrus. We shall return to some of his textual proposals later.
We are both grateful and gratified that J. considers many of our own proposals excellent and thankful for his remarks about the complete reliability of our readings. We also freely accept his charge that our text is conservative. He further writes, however, that we have refused “to admit any emendations, even where the papyrus is blatantly ungrammatical or nonsensical”. He lists a number of such instances, but nearly all succeed only in showing his propensity for hypercorrection. The following are all the cases he enumerates (the vast majority of them pertain to emendations he has advanced in the last decade and which we chose not to adopt in our edition):
Col. 9,9: “a plural antecedent [i.e. ὅσα ] governs a singular participle [i.e. ἐπικρατηθέν ]”. The lack of agreement is easily resolved by writing ἐπικρατηθέντα, as J. has proposed, but actually nothing need be changed. J. may not have noticed our comment on this passage, where we acknowledge the syntactical incongruity but point to an exact parallel in Euripides, H.F. 195-197. The construction is “eine ganz gewöhnliche Erscheinung”, as Wilamowitz calls it in his commentary on the play. One may further consult, inter alia, Bond’s comments, in his edition of the same play, about a generalizing plural commonly followed by a typical singular, and Maas, Textual Criticism, p. 37. In our passage the presence of the Attic construction makes the case more complicated, or less easily recognizable.
Col. 15,7-8: “with better word-division and punctuation”, as he writes, J. makes the Derveni author state that “(Orpheus) means that his [=?] rule has existed from the point when (Mind) became king.” Frankly we do not understand this. In this passage the Derveni author is attempting to explain what Orpheus means by saying that “Kronos” followed “Ouranos”, not to estimate the duration of Mind’s rule. Obviously a kingly rule begins when someone becomes king; but can J. indicate any time when Mind/Air, according to the author’s conception of the history of the universe, was not king, i.e. absolute controller of all other elements, and his rule did not yet exist?
Col. 16,10: “auton lacks the article”. J. is correcting one of the older versions of our text. In the edition under review there is no such word.
Col. 21.13: “‘he was named to have been born’ is ridiculous”. Indeed it is. But this is J.’s understanding of the syntax and his translation of the text, not ours. Our translation is to be found on p. 137, and our notes on pp. 252-253. In proposing his easy emendation, J. overlooks the fact that in this column the author is interested in names, not in what people think about the birth of gods or the creation of the present world.
Col. 24,4: J. is right. We should have been more consistent in our treatment of ε for [e:] and written ε < ι > (cf. our comments on p. 10, where we mention the word in question), as we did in cols. 15,1 and 20,8.
Col. 24,6: there is certainly no “lack of governing verb”, as J. asserts. Our punctuation and translation make it quite plain that we take φαίνειν αὐτήν as standing in apposition to the preceding τοῦτο, and in such a case a governing verb would be wrong.
Col. 25,1: we do not understand J.’s remark that “a superlative is needed instead of a comparative”. How can we make λευκότατα more superlative? It is very likely, however, that J. meant to write that a comparative is needed instead of a superlative. In his edition he corrects the second word of the line from λαμπρότητα to λαμπρότατα (perhaps thinking that we had misread the papyrus in an early working text of ours printed, without permission, by Merkelbach in ZPE 47, 1982), and then, assuming an “error perfacilis” caused by his new text, further emends the following λευκότατα to λευκότερα, because he is troubled by τῶν ἄλλων. All this is unnecessary. The text, as the papyrus has it and as we print it, gives perfect sense; λευκότατα τῶν ἄλλων, “whitest among elements”, is correct Greek and parallels of it abound; and Burkert’s suggestion (cited in our note ad loc.) gives a good idea as to what the text before the beginning of the column might have been.
Col. 25,2: “kai has dropped out before kata”. A καί may, but certainly need not, have dropped out. The reader may, if she or he wishes, place a comma before κατά, as we have done in the translation.
Col. 26,11: “a dative [i.e. αὐτῶι ] agrees with an accusative [i.e. παρακλίναντα ]”, which means that we should have adopted J.’s “essential”, as he calls it elsewhere, correction and written παρακλίναντι. H. W. Smyth, Greek Grammar, 1062, to cite only an elementary handbook, notes that a predicate participle referring to a dative may stand in the accusative in agreement with the unexpressed subject of the infinitive. Examples may be found in Lysias, Xenophon, Isaeus, and others.
We are further accused of retaining “a number of supplements that have been improved upon by recent scholarship”. J. is quite entitled to consider his, and Betegh’s, θέσιν at col. 7,4 (the only example he offers) an improvement, as we are entitled to differ; perhaps we should have stated our reasons for not adopting his interpretation of the passage, and in this he is probably right. But he is certainly wrong in believing that κεἰ in the next line “deprives the sentence of a main verb”, perhaps thinking that we mean “and if”. Our translation shows that we mean “though” (cf. Denniston, Greek Particles, pp. 301-302); the supplement may be wrong, but the sentence does have a main verb. Finally, in the admittedly troublesome passage at col. 8,9-10 our translation indicates that, contrary to J.’s accusation, we take ἐπικρατεῖ as the verb of a main, not a subordinate, clause.
As for the translation, whose only aim is to demonstrate how the editors understand their text, J. states that it “accurately renders the Greek”, but he immediately afterwards remarks that “there is much[!] that makes little sense”; this is due, of course, to our not having accepted his emendations and supplements. He cites as an example our translation of col. 12,3-10, where he believes that “the logic and sequence seem crazy”, adding in a footnote that he “cannot make head or tail of the logic of the passage” as explained in the commentary. (Others, Betegh and Jourdan among them, had no difficulty with our supplement.) Our commentary seems clear enough, but let us be more detailed. In this passage the author wishes to prove that, in the Orphic verse which he cites in col. 12,2, “Olympus” = “time”. He does not take into consideration that “Olympus” may refer to a mountain but endeavors to prove that it cannot mean “sky”, as some think. His argument runs as follows: (a) One may not say that the sky is long, though one may say that it is broad (J.’s translations “longer” and “broader” are wrong; cf. Smyth, 1080, on double comparison). (b) One may say that time is long. (c) Orpheus uses the expression “broad sky” when talking about the sky, and the expression “long Olympus” when talking about “Olympus”. (d) Since the epithet “long” can only be applied to time and to Olympus (the expression “long sky” being impossible), by saying “Olympus” Orpheus can only mean “time”, i.e. “[long] Olympus” = “[long] time”. To put it inversely, by saying “Olympus” Orpheus cannot mean “sky”, for thus we would be confronted with the equation “[broad] sky” (a possible expression) = “[broad] Olympus” (an impossible expression). Adopting J.’s supplement and translation (if we understand the latter correctly), we must needs then saddle Orpheus with the formula “broad Olympus”, which no epic poet except Quintus Smyrnaeus ever used. The reasoning employed by the Derveni author is similar to that used by ancient scholiasts to prove that in some Homeric passages (cf. the scholia on Iliad 1,402 and 15,21) “long Olympus” must refer to a mountain and cannot refer to the sky, since the latter is called “broad”, i.e. “broad sky”, not “broad Olympus”. J. is quite right in stating that misidentification of the subject in col. 17 may cause grave problems. He disagrees with our choice of subjects and offers a translation which contains the startling sentence “Why (Zeus) was called ‘Air’ has been revealed earlier”. It would indeed be a revelation, had Orpheus (or anyone else, for that matter) called Zeus “Air”. (J. has momentarily forgotten that the exact opposite holds true.) It is far more logical to translate “why (air) was called air”, and we have a long discussion on this, pp. 218-219.
It is a pity that J. chose to focus his attention on our refusal to adopt his text and to follow his quite readable but very free and occasionally inaccurate translation. There are important things he might have written about our understanding, or misunderstanding, of some difficult, crucial passages, and his views would have been most welcome, as he is a scholar who has worked long and pondered deeply upon the meaning of the text, and whose ideas have often caused us to reexamine ours.
A parting word. We never dreamed of our edition as being “the definitive edition” that J. wishes we had produced. Can he point, however, to a “definitive edition” of a fragmentary papyrus text — or of any text for that matter? There will be other editions after ours, and better ones at that; otherwise there would be no point to their appearance. But will one of them be “the definitive edition”, except perhaps in the eyes of its creator? Our goal was more modest. We are grateful that J. considers our book, in one perspective at least, “a gold-mine”, and we are satisfied with such an assessment.
[For a response to this response by Richard Janko, please see 2006.11.20.]