BMCR 2005.09.63

Chariton. De Callirhoe narrationes amatoriae. Bibliotheca Teubneriana

, , De Callirhoe narrationes amatoriae. Bibliotheca scriptorum Graecorum et Romanorum Teubneriana. München/Leipzig: K.G. Saur, 2004. xxi, 150 pages ; 21 cm.. ISBN 3598712774 €80.00.

Chariton’s first- or second-century romance has had a tough time. It survives in only one thirteenth-century manuscript (F = codex Florentinus Laurentianus Conv. Soppr. 627, which was rather negligently produced), four short papyrus fragments, and a partial transcription by Wilcken of a sixth- or seventh-century palimpsest codex with an inferior text. Moreover, until Blake’s edition (Oxford 1938) not one of Chariton’s editors ever saw the manuscript!1 Blake’s was from its appearance the most authoritative text until recently. It has its faults, to be sure, some of them serious, but the most pressing are simply that it is difficult to obtain, at least for private libraries and the collections of smaller classics departments, and that so much has happened (critical work, reappraisals of the genre, papyrus discoveries) in the decades after its appearance. As scholarly interest in the Greek novel rose, there were three major attempts to replace Blake. The first, an edition by F. Zimmermann, never appeared. The second, a projected Loeb volume by Reardon himself, was cancelled due to budgetary constraints. The third, G. Molinié’s Budé (Paris 1979), did appear, but the text was a disaster. In the end a Loeb was produced in 1995, edited and translated by series editor G. P. Goold, who gave us both a clear translation and a readable and accurate text — by far the best up to that point. But although Goold applied considerable energy and intelligence in editing his text, without a full apparatus and with no pretensions of being a critical edition this volume was nothing more than a stopgap for scholars who wished to work on the text of Chariton (though it will have lasting value on its own terms). Now Reardon’s new edition in the Bibliotheca Teubneriana, the latest product of his decades-long engagement with this author, will, like Blake’s text in its time, make all earlier editions essentially obsolete. Its virtues are many; its defects are few and relatively minor. We will, of course, continue to debate Reardon’s individual editorial decisions and turn up errors here and there, but no one will doubt the authority and utility of this volume. Scholars have had to wait decades for a well-edited critical edition of Chariton, but the wait has been worth it.

Without biographical details or other writings of Chariton to discuss, Reardon can give us an economical (12 pages) praefatio that covers dating (Neronian for Reardon, who takes the Callirhoen at Persius 1.134 to refer to our novel), accurately summarizes the transmission of the text and the production of earlier editions, and remarks briefly on Chariton’s language and his own approach as an editor. One might wish for more, but Reardon gives pointers to the important bibliography, and his own views on Chariton are widely available elsewhere. It is surely the text and apparatus that really matter here.

At first inspection, Reardon’s text seems very much like a slight reworking of Blake’s. Because new paragraphs are generally begun in the same place and the punctuation only rarely differs, a certain physical resemblance reinforces this impression. A closer look, however, reveals quite another picture. Reardon has thoroughly reconsidered almost every editorial decision made by Blake and practically every conjecture ever made to Chariton.2 If Chariton’s text does not lend itself to massive editorial interventions, there are nevertheless ample opportunities to correct the dozens of places where F very clearly presents a bad text and to make slight conjectures in the dozens more where any given critic can see a possible need. Asking, for instance, whether or not Reardon’s text is conservative is not really useful.3 Sometimes, as with the five short lacunae he prints without supplement in 1.1.5-6 (versus one in Blake and Goold and none in Molinié), he refuses to jump too quickly at a conjecture simply because a conjecture exists. Likewise, he is far more likely to obelize deeply corrupt passages and relegate solutions to the apparatus than print a conjecture that is not wholly satisfactory or was made exempli gratia. This is conservative, but more importantly it is sound judgment. But throughout the edition, using just as sound a judgment, he readily prints a text that varies at many points from F but without Blake’s tendency to emend for no reason at all. An editor of Chariton must bring not dogmatic editorial theory to the task, but, in Reardon’s own words, he must “surtout, apporter à l’établissement du texte une bonne connaissance du langage grec post-classique et un jugement meilleur que celui de Blake.”4 I think it can safely be said that Reardon has done just that.

To give some idea of the changes involved in the text, in book 1 (pp. 1-22), there are only two pages on which Reardon’s text is identical to Blake’s (p. 1 [1.1.1-3] and p. 11 [1.6.4-1.7.4]). On the other end of the spectrum is page 22 (1.14.7-10), where there are the most divergences in book 1. I count some 9 differences (aside from punctuation): twice Reardon accepts a change in word order to avoid hiatus (one from Reeve CQ 21 (1971) 514-539, and one Reeve per lit.); twice he inserts the pronoun με for clarity; he accepts D’Orville’s διώκεις for F’s γινώσκεις; he twice follows Cobet, changing ἠνέχθην to ἤχθην, and deleting the otiose αὐτῇ in the last sentence of the book; and he prints τοιούτῳ διαζευχθεῖσα πάθει where Blake printed τοσούτῳ διαζευχθεῖσα πελάγει ( τοι – is Reiske’s conjecture for F’s τος is Hercher’s conjecture for F’s παθ; each of the earlier scholars solving the same problem in a different way, but Reardon chooses to follow Reiske and print the manuscript’s noun — rightly, I think, despite the perfectly good sense of τος, because τος is an easy mistake for τοι, though one never knows with the scribe of F). The upshot of this is that Blake’s text here varies in 4 places from F’s readings, while Reardon’s does so in 10 places. Conservative? Certainly not, but perfectly intelligible. One can see how Reardon allows room for his own insights while incorporating work since Blake and giving proper place to scholarship stretching back to the beginning of the modern study of Chariton.5

Still, Reardon’s approach allows inexplicable inconsistency to creep in. For instance, while he resists absolute standardization of Chariton’s language,6 everywhere F gives a form of φιλόνεικος (so spelled), he standardizes to φιλόνικος. This may be etymologically correct, but I suspect that by Chariton’s time the former spelling predominated. Reardon rightly approaches Heibges’ work on clausulae7 (cf. praef. xiv) cautiously where textual matters are above the level of minutiae,8 but he follows Heibges’ lead on which form of the 2nd- and 3rd-person reflexive pronouns to print to produce favored clausulae. For example, on p. 26 Reardon prints uncontracted σεαυτῆς for F’s σαυτῆς in 2.2.5 but just below in 2.2.6 he contracts F’s ἑαυτὴν into αὑτὴν as Heibges suggests. Such changes — harmless as they are in the greater scheme of things — only make rare clausulae rarer and favored clausulae more favored and seem to me to be based on work that is in serious need of an updated treatment. Put simply, prose rhythm is a far less reliable guide to what Chariton wrote, than, say, the avoidance of hiatus, at least on our current understanding. Another matter entirely, and one far less typical: at 2.6.5 he emends ἠπείγετο πρὸς τοῦτο to ἐπείγετο περὶ τοῦτο based on 3.6.2 περὶ τὴν ἀνάληψιν ἠπείγοντο. But at 3.6.2 he accepts and prints ἐγίνοντο, not ἠπείγοντο.

As with any text, even when something is not amiss with the editor’s method, one will find quibbles. To take a few examples at random, I cannot see how D’Orville’s supplement of < καὶ ῤοδογούνην τῷ ἀνδρί >, which Reardon prints at 8.3.2, can possibly be right, since it is the Persian queen who matters at this point, not her handmaiden. I also do not think that Chariton is likely to have written, as Cobet suggests and Reardon prints, modal ἄν as the first word of the resumed quotation after εἶπεν at 5.10.6 (surely it belongs after ἀπέθνησκον if it belongs at all). At 4.1.7 it may well be right to print ὁ Λυδίας < καὶ Ἰωνίας > to match the form of the other mentions of Pharnaces’ satrapy in the novel, but Arrian at least offers a parallel for both forms of the phrase in a single work without distinction (cf. An. 1.12.8, 1.16.3, and, showing the clear priority of Lydia, 1.17.7). Even a good edition of a text of this length cannot please everyone in every place. The real question is whether as a whole it is an improvement over previous editions. The answer in this case is yes, without any doubt whatsoever, and a major improvement at that.

That high quality can also be seen in Reardon’s apparatus. Blake’s double apparatus had the admirable goal of keeping the most important information readily available in the top apparatus while recording every conjecture in the lower, but the result was often unwieldy. Reardon has been much more selective, and one can hardly fault him for not wishing to record most of the obviously incorrect conjectures that have piled up around Chariton’s text. One can always check Blake, of course, for the older conjectures, but here and there Reardon might have been a bit more generous, perhaps especially in the case of Blake’s own conjectures — though often clearly wrong, they have been the basis of translations and scholarly work over the last half-century and more.9 Reardon’s apparatus is also clearer and more accurate than Blake’s, both in attributing emendations and in reporting readings.10 His reporting of the codex Thebanus and the papyrus fragments is conscientious, and it is a great pleasure to have the post-Blake discoveries now incorporated into an edition. There is a small apparatus fontium in which Reardon adds a few sources previously unknown to me, but I hope that the appearance of this edition will give rise to further work in this area. I suspect that several citations yet lurk in the text of Chariton, and despite Papanikolaou’s (Göttingen 1973) lists of citations, themselves mostly derived from Cobet and others, there is more work to be done.

On the whole, then, this is an excellent edition of a text that has long been in need of an excellent edition. Bryan Reardon was just the person for the job, and he has done it very well indeed. Every scholar with an interest in Chariton needs this edition. No other will do. Barring some further discoveries in the sands of Egypt, that situation is unlikely to change for some decades.

That said, I hope that I can end with a few additional criticisms without detracting from that judgment. First, I would have welcomed an index verborum, though to produce it would have been a huge task, and more and more scholars rely nowadays on computerized searches rather than a printed index. One can contrast Reeve’s Longus Teubneriana, which has a tremendously useful index verborum. Second, the price is too high. Third, Reardon’s truly impressive achievement is marred by Saur’s having let far too many misprints find their way into the text. To have a few is expected. To have them in the praefatio and apparatus is unfortunate.11 To have so many in the text is absurd. Even though almost all of these are quite minor, they are irksome, and I list those I found.

References are to Reardon’s pages and line numbers (e.g., 12.320 = page twelve, line 320 of book 1):

4.68 ᾔδει, not ᾖδει

9.232 δικαστηρίῳ, not διακα

21.573 μήν needs a grave or a comma after it

30.175 Ἔρως, not Ἕρως

32.222 σιωπῆς, not σιωτῆς

40.473 καί should have a grave accent

42.11 a comma is needed after φησιν before the vocative

49.234 εἰμί, not εἰμὶ

56.434 ἔπεισε, not ἕπεισε

57.462 ἔτεκε, not ἕτεκε

65.77 ἔρωτι surely needs to be capitalized here

78.4 πολυτελῶς, not πολι

79.46 παρεσκεύασεν, not παρεσκεύσεν

80.73 αἰφνίδιον, not αἰφνίδιοιν

82.142 Περσίς should have a grave accent

82.143 Ἕλληνες, not Ἔλλ

84.176 ὑβρισθῆναι, not ὕβρισθῆναι

87.287 ἐμὸς has been omitted after ξένος

92.419 συνεβούλευον, not συνεβούλενον

95.513 ἀπηλλαττόμην, not ἀπηλλαττόμεν

103.196 ἀστάθμητον, not ἀσταθμήτον

105.242 ἐκείνῳ, not ἐκείνῳι

123.302 ἐνεπλήσθη, not ἐνεπλήθη

125.351 period needed after βουλομένη, not comma

130.124 Κύπρου not Κύπρον

131.165 ἑκάστου, not ἑκάστον

131.171f. ἀνέκραγεν is missing its accent.

136.318 ἔδειξεν, not ἔδειξὲν

140.408 ἐπολιτεύετο, not ἐπολιτέυετο

In addition to these, Reardon sent an erratum sheet of his own via the press to reviewers. It indicates that at 125.348 the text should read οὐδεμιᾷ γέγονε κακόν and that the apparatus ad loc. ought to read: οὐδεμιᾷ Cob. : οὐδεμία φ κακὸν Cob. (per compendium; vix dispicitur) : κακὴ dubitanter Apogr. : κάκη Rei., sed nusquam alibi reperitur ap. Char.


1. W. E. Blake, “The Overtrustful Editors of Chariton,” TAPA 62 (1931) 68-77.

2. For those unfamiliar with Chariton, the amount of text-critical work applied to this author is enormous. Just in the year after Blake’s edition, one had short articles on textual matters by W. B. Stanford ( Hermathena 51 [1938] 135-139), H. J. Rose ( CQ 33 [1939] 30), and W. Morel ( CQ 33 [1939] 212-213). The most recent publication concerned with the text of Chariton is A. Borgogno, “Per il testo di Caritone di Afrodisia,” Prometheus 30 (2004) 246-252, which appeared just before Reardon’s edition appeared, too late to be used in its preparation.

3. One can instructively read R. Hunter’s review of Goold’s Loeb (BMCR 96.02.09) alongside J. Birchall’s ( Scholia Reviews ns 5 [1996] 25): The former writes: “this is not a conservative text.” The latter: “The text itself is relatively conservative.”

4. In his devastating review of Molinié’s Budé at REG 95 (1982) 157-173 at 158.

5. With greater circumspection than Goold in his Loeb, Reardon also incorporates some of Jackson’s unpublished marginalia into his text and apparatus. Similarly, all those who use this edition will be grateful that Michael Reeve and Robert Renehan read through the whole. Their conjectures made through correspondence with Reardon are recorded in the apparatus (and Reeve’s marked in the apparatus with asterisks, as are Jackson’s marginalia).

6. With good reason: tam diversae erant linguae Graecae formae ut non facile scire possis ad quam normam exigendus sit auctoris nostri sermo (xv).

7. S. Heibges, De clausulis Charitoneis (Halle 1911).

8. For example, in 1.14.9, discussed above, where Reardon retains F’s πάθει, Heibges very much prefers Hercher’s πελάγει, not only for sense, but especially because it produces a much more favored clausula.

9. For instance, Reardon translated Blake’s ἔπεισεν at 6.1.2 in Collected Ancient Greek Novels (Berkeley 1989), but the reading is not found in his apparatus.

10. Of the several entries in the apparatus I had occasion to follow up on, I noted only two that were problematic: at 1.1.9, where Reardon prints Blake’s οὐδὲ πειρᾶσθαι, he reports F’s reading as οὐδὲ πειράσαι, but as far as I know (and so Blake reports), the acute is Cobet’s correction on moral grounds of F’s circumflex (Cobet [1859] 250). At 1.2.3 ἐτάκημεν is attributed to Blake by Reardon. Blake, by contrast, attributes it to Beck (to whose edition I have no access). Presumably, there are other such issues, but on the whole Reardon is a distinct improvement.

11. I did not systematically collect typographical errors in these contexts, but I note a few here: Cobet’s Annotationes criticae are at 229-303 in Mnem. 8, not 229-309 (xvii); ζανκογιάννης, not ζαγγ (xxi); 2.3.5, not 2.3.10 (?) (p. 29 app. ad 163); συνεκάλεσε, not συνέκαλεσε (p. 61 app. ad 594); καλλιρρόην, not καλλιρόην (p. 78 app ad 1); ἐλείβετο, not ἔλείβετο (p. 80 app. ad 76).