Editing Chariton is not a straightforward business. His novel in its entirety (albeit with some significant lacunae) is preserved only in a thirteenth-century codex now in the Laurentian Library in Florence (F = Conv.Soppr.627), which is also the sole witness for Xenophon of Ephesus and the only manuscript of Longus to supply a large chunk of the first book. Our knowledge is supplemented by three papyrus fragments of similar date (mid- to late-second century ce), not too long after the composition of the novel, and by the bizarre case of the so-called codex Thebanus deperditus (designated W by S.M.), a fragment of a seventh-century parchment palimpsest, provisionally transcribed by U. Wilcken, but destroyed in a fire before it could be properly edited. The papyri present a text not dissimilar to that of F, although there is no way of knowing whether they originally did so over the whole length of the novel or how they are related to one another and F's earliest ancestors. W, however, differs significantly from F, offering a somewhat fuller version of 8.5.9–7.3, either (as Wilcken argued) because F represents an abbreviated version of Chariton's original or (as argued by Zimmermann, followed by S.M.) because W's text has been expanded to make it more attractive to sentimental readers. That expansion presumably occurred after the date of the papyri, but how long after is anyone's guess. However, the possibility of such manipulation clearly calls into question the stability of the novel's transmission, and the exact status of its one medieval witness. A parallel case occurs with Achilles Tatius, where papyri offer a reordered version of an important episode. Although not exactly textes fluides, the novels were apparently subjected to more scribal creativity than more canonical texts.

The main problem for an editor is the abject quality of F. In Longus, where F can be compared to a better manuscript, the variants are sufficient in quantity and quality to have given rise to a hypothesis of different authorial recensions, though in truth they are due to F's ineptitude. In the case of Chariton, comparison with the papyri suggests that F commits an error in every other line of the text. Editors, from the editio princeps of J.P. d'Orville (1750) onwards, have exercised their judgement to make sense of F's offering. The text has been improved by a host of corrections, ranging from the obvious to the brilliant: of particular note is the work of great nineteenth-century philologists such as R. Hercher, S.A. Naber and C.G. Cobet, many of whose conjectures were subsequently confirmed by the papyri, and, in the first half of the twentieth century, J. Jackson. Scholars continue to chip away. One thing is certain: a conservative text of Chariton is a non-starter.

The first fully critical edition was that of W.E. Blake (1938). G. Molinié's Budé of 1979 was deeply flawed (see the devastating review by B.P. Reardon, REG 95 [1982], 157–73). Reardon's own Teubner (2004) has been the standard text since its publication, while G. Goold's Loeb (1995) offers a worthwhile and thoughtfully edited text, but with a minimal apparatus. So what does S.M. bring to the table? As he says (p. xvi), he can take note of work that has appeared since Reardon went to press. In addition, he offers a more comprehensive apparatus, in order to give at least a cameo role to conjectures that were at ‘risk of falling unjustly into oblivion’. More unconventionally, at the end of the book is an apparatus criticus additicius (pp. 167–79), which aims to record every conjecture ever made, with the exception of those clearly based on incorrect information about F's reading. No doubt there is gold to be found among the dross of League Two conjectures for those with sufficient patience and store of aspirins to sift through these densely printed and inconveniently located pages, but their main utility will be to enable someone with a thought about the text to check quickly whether anyone has ever had it before; however, such a resource begins to be out of date the moment a new conjecture is published. Finally, on each page of the text, above the apparatus criticus, is an apparatus fontium, identifying the source of Chariton's many allusions, and in some cases outlining (in Latin) the original context. Reardon already offered something of the kind, but S.M. is fuller, and the contextualisation is useful. The danger, as S. Trzaskoma has recently argued (‘Citations of Xenophon in Chariton’, in K. Chew, J.R. Morgan & S.M. Trzaskoma [edd.], Literary Currents and Romantic Forms. Essays in Memory of Bryan Reardon [2018], pp. 65–79), is that it is all too easy to base interpretations of Chariton's intertextuality on the assumption that all the allusions have been spotted, whereas there are still plenty of less obvious ones to be found, particularly to the works of Xenophon. However, the real point of this new edition is that it will ultimately work in tandem with the commentary on Chariton by S.M. and M. Baumbach, whose publication it precedes by a few months, and which will key its lemmata to S.M.'s text (and, no doubt, elaborate on some of the editorial choices).

S.M.'s text tends to the conservative. I noted more than 150 places where he retains F's reading, while Reardon emends it. Sometimes there is a clear ideology behind this choice. For example, having determined the status of W, S.M. systematically rejects its more expansive variants, while Reardon is prepared to consider each on its merits. He also seems unconvinced that Chariton avoids hiatus, regularly retaining F where a simple transposition would eliminate it (as at 1.12.7: τέθνηκε γὰρ αὐτοῦ ἡ γυνὴ ἧς ἤρα F, S.M. : ἡ γυνὴ γὰρ αὐτοῦ τέθνηκεν ἧς ἤρα Reeve, Reardon). The argument, I suppose, is that an editor should not impose stylistic preferences on a text that gives satisfactory sense; but Reeve's famous article on hiatus in the novelists (CQ 21 [1971], 514–39) presents a powerful case to the contrary. Sometimes, however, S.M.'s allegiance to F is less easy to understand; at 5.3.7, for instance, Dionysios leaps out of his horse (πηδήσας ἐκ τοῦ ἵππου F, S.M. : πηδήσας ἀπὸ τοῦ ἵππου Cobet, Reardon); or at 6.5.7 the eunuch Artaxates advises Kallirhoe to sting in order to please the Persian king (νύττε ὅπως ἀρέσῃς μᾶλλον αὐτῷ F, S.M. : ἄνυτε κτλ. Goold, Reardon). Other cases are more arguable, but there is a significant number of places where, it seems to me, Reardon's more interventionist approach results in a preferable text, though, to be fair, the corrections are invariably to be found in S.M.'s apparatus. The traffic is not all one way: there are places where Reardon retains F and S.M. corrects, and a larger number of places where the two editors agree that change is necessary but opt for different solutions. Finally, S.M. has, over recent years, published several papers on the text of Chariton, and some of his own conjectures find their way into the text of his edition. Some are just a matter of good housekeeping, and a few do not convince (for example at 1.1.12 the infinitive νυμφαγωγῆσαι, for F's νυμφαγωγοῦσαι, dependent on παρῆσαν, is not supported by the parallels cited in the apparatus; at 6.7.7 the infinitive depends on a different verb, and in any case F's text needs emendation; at 8.3.5 impersonal πάρεστι with infinitive is something quite different; and at 5.2.2 Reeve's insertion of θέλων to govern the infinitive seems a better solution). However, there are worthwhile new readings proposed at 1.7.1 (ἐκ Λυκίας for ἐξ ἀδικίας); 3.10.1 (συνεκάλεσε πιστοτάτους for συγκαλέσας πεισθέντα, obelised by Reardon; 6.7.7 (<οὐκ> ἀκούεις;); 7.6.10 (κλοπή for κακή, obelised by Reardon), and elsewhere.

The book is immaculately produced, with an attractive Greek font and next to no typos. To sum up: this is a serious work of scholarship, though it does not mark an unqualified advance on Reardon. My sense is that those who are interested in Chariton will continue to use Reardon, while those who are interested in the text of Chariton may well prefer S.M.