Michel Patillon (ed.), Corpus rhetoricum. Tome II, Hermogène: les états de cause. Collection des universités de France. Série grecque, 470. Paris: Les Belles lettres, 2009. Pp. cvii, 205. ISBN 9782251005539. €31.00 (pb).
Reviewed by Malcolm Heath, University of Leeds
Michel Patillon, who has produced a series of important editions of Greek rhetorical texts in the Budé series,1 is currently engaged in a project to edit the rhetorical corpus which accreted in late antiquity around Aphthonius' Progymnasmata and four works attributed (two falsely) to Hermogenes. The first volume (published in 2008, and reviewed in BMCR 2008.09.17) included an anonymous introduction to rhetoric (number 4 in Rabe's Prolegomenon Sylloge) and Aphthonius; the pseudo-Hermogenean Progymnasmata (which was never part of the corpus) was included as an annex. This second volume is devoted to Hermogenes' περὶ στάσεων (On Issues). It provides us with the best text and critical apparatus to date, along with a lucid French translation,2 a valuable introduction and extensive notes.
Rabe's Teubner has been the standard edition since its publication in 1913 -- a fine piece of work, though Rabe himself made no secret of its limitations.3 Patillon systematically reports the readings of a wider range of manuscripts than Rabe (he helpfully retains Rabe's sigla: but for detailed descriptions of the manuscripts, and for a justification of the stemma printed on p. lxxxvi, one must consult volume 1). In constituting the text Patillon corrects Rabe's bias towards the manuscripts Pa and Pc. This has yielded improvements in the text in many places, though they are mainly minor. Patillon also pays close attention to the indirect tradition -- that is, to the abundant ancient commentaries. Readers may feel that the principle consequence is an apparatus cluttered with evidence of how carelessly the text was handled in the indirect tradition. But amid all the flotsam and jetsam, things of genuine value come to light. At 1.25.8, where Rabe (34.22-35.1) printed ἐπιδεικτικῶς τὸ ἐπιδεικτικὸν, Patillon adopts πανηγυρικῶς τὰ ἔνδοχα from Sopater. The manuscripts are hopelessly divided here, but πανηγυρικῶς (or τὸ πανηγυρικὸν) occurs widely across the direct tradition, and is consistent with Hermogenes' usage in his work on types of style (in which ἐπιδεικτικῶς and cognates do not occur); and τὰ ἔνδοχα is an improbable corruption. Patillon is right to conclude that Sopater guides us to the truth.
Another significant improvement can be found at 10.18.3f. Where Rabe (87.20-88.1) had ἢ· εἰ μηδὲ τοῦ πατρός, ἀλλ' οὐδ' ἑτέρου τινός, Patillon prints ὁ δέ· ἢ τοῦ πατρός ἐστιν, ἀλλ' οὐχ ἑτέρου τινός . The case under discussion is a conflict of law: a disinherited son, who has remained on board an abandoned ship belonging to his father, claims ownership under a law of salvage; his claim is contested under a law debarring disinherited sons from their fathers' property. Under the head of forcible definition (biaios horos, 'définition violente') the son suggests that the ship is not his father's property, but is the storm's possession (so that there is no legal obstacle to his own claim). According to Rabe's text, an alternative forcible definition is that the ship does not belong to the father, nor to anyone else. That is how I understood it in my translation and commentary (which, as I now see, was rather uncritical in its adherence to Rabe).4 If, abandoning Rabe's preference for Pa and Pc, we adopt the variant ὁ δὲ, we no longer have an alternative forcible definition, but the opposite party's reply (cf. ὁ δὲ at 10.15.9 = 87.1 Rabe): the ship belongs to the father, not to anyone else. This gives better sense in respect of rhetorical theory, and is supported by the commentators' exegesis.5
None of this, however, lends support to Patillon's ἢ, which is (though the complexity of the corruption in this passage obscures the fact) an emendation. In the latter part of the text, Hermogenes has a habit (according to the transmitted text) of introducing telegraphic indications of how a head of argument might be formulated in a given case with the word εἰ; Patillon persistently changes this to ἢ (nine times in all).6 The error is, of course, one that could easily occur in any individual passage. But Patillon's text entails that the corruption has occurred systematically and across the whole tradition (the manuscript tradition does not preserve ἢ anywhere in this text in a context of this kind); that is not so plausible. The idiom is not particularly difficult to understand (the speaker is envisaged as posing the question whether...), and the commentators had no qualms about it. Patillon's policy displays a misplaced fussiness -- something that has manifested itself in his editions of other rhetorical texts.7
This fussiness also leads him to mark lacunae in two passages. At 5.6.1 (= 66.7 Rabe) he is worried by the omission of a head of argument: the text proceeds from definition to assimilation, without the counterdefinition that is both a routine part of the sequence of heads introduced by definition and a logical stepping-stone to assimilation. Patillon dismisses the explanations offered by the commentators as 'autant d'arguties'; I remain broadly in agreement with Sopater's analysis (RG 5.168.26-169.2, cross-referring to 154.9-20: see my commentary on 60.8-14 and 66.7-12 Rabe). At 12.9.2 (= 92.3 Rabe), in a case of ambiguity in the wording of a law regulating prostitutes, the prosecutor's use of the common topic against prostitutes is to be countered by an objection (metalepsis). Patillon claims that what follows Hermogenes' announcement of the objection does not correspond to that head of argument. The ancient commentators, who knew perfectly well what metalepsis was, did not see a problem. They were right. The defendant's conduct cannot be judged by the standards that govern free women, because of the constraints of her profession; the common topic is therefore irrelevant to resolving the legal ambiguity. This is a straightforward instance of objection based on person (see my commentary on 91.15-92.2 and 92.2-7 Rabe).
Patillon's treatment of the commentators is in general more careful and more respectful. It is not only that he takes them seriously as witnesses to the text. His notes draw extensively and very effectively on the evidence they provide for an ongoing (and intelligent) tradition of professional debate about the interpretation of Hermogenes' text, and about the substantive issues of rhetorical theory and practice that arise from it. Anyone needing an orientation to this difficult and confusing body of texts would do well to read the brief account of the commentators in the introduction. I am delighted to discover that we have reached similar conclusions on the character of Sopater's commentary (pp. lxviii-lxx). I am not surprised, given the complexity of the material, that we do not agree on everything: I think it is a mistake to describe the Sopater material incorporated in the composite commentary printed in volume 4 of Walz's Rhetores Graeci as coming from 'une édition plus complète' of the version printed in volume 5 (p. xxix n.4): it is, at the very least, based on a thoroughly reworked version of the commentary of which RG 5 is an abridged version.8
Now that I have begun to touch on details of the introduction and notes, readers impatient of the technicalities of late antique rhetorical theory and the source-critical complexities of its reconstruction may prefer to skip to my concluding paragraph.
Here, as in previous publications, Patillon worries about the identification of the author of this treatise with the Hermogenes in Philostratus' Lives of the Sophists (2.7, 577). I see no reason why Philostratus' account of Hermogenes' career as a declaimer should be expected to reflect, or to throw light on, the composition and reception of the technical treatises; it should not be forgotten, either, that Philostratus' account is highly polemical and tendentious, though the polemic is not so extreme as one might infer from Patillon's 'prématurément sénile'. Patillon (p. x) argues that the account of Marcus Aurelius' visit to Athens in Dio (71.1) refers to 'exposés techniques' by Hermogenes; but that puts too much pressure on an account that is marked as hearsay (λέγεται γὰρ... ), and may not be precise. Though it is credible that the emperor still hoped to learn something from the distinguished philosopher Sextus (Philostratus Lives of the Sophists 2.1, 557), it is much harder to imagine him expecting to profit from listening to technical lectures on rhetoric (he was in his mid-50s at the time of this visit). Philostratus' account is more credible: the emperor visited a school of rhetoric, and enjoyed the opportunity to hear a star pupil performing. Philostratus is maliciously interested in the star pupil's subsequently failure to live up to his early promise as a declaimer; whether he went on to write technical works was completely irrelevant to the point of the anecdote.9
The Suda entry on Hermogenes (Ε3046) records that 'the philosopher Musonius' was his pupil. Patillon is right to say (p. xvi n.2) that the identification of this Musonius is uncertain. But it is strange that he gives a wholly irrelevant reference to Kurt von Fritz's Pauly-Wissowa article on Musonius (1) (that is, the first-century Musonius Rufus), and does not mention Musonius (17), a Stoic Musonius contemporary of Longinus (F4 Patillon-Brisson = Porphyry Life of Plotinus 20), who is a chronologically plausible candidate.10
Patillon proposes (p. viii n.2) a repunctuation of the Suda article on Menander Rhetor (Μ590) that would modify his bibliography: 'He wrote a commentary on Hermogenes' Art and Minucianus' Progymnasmata, etc' becomes 'He wrote a commentary on the Arts of Hermogenes and Minucianus, Progymnasmata, etc'. This is unconvincing and uneconomical: there is independent evidence for Minucianus' Progymnasmata (Suda Μ1087, admittedly complicated by the conflation of Minucianus with a later homonym), but we have no independent evidence that Menander wrote progymnasmata or commented on Minucianus' treatise on issue-theory.11
There is sometimes a curious failure of integration in Patillon's treatment of source material. For example, on p. 99 he asserts that there were only four issues according to Lollianus (in the early second century AD). He makes the same claim on p. xlvi, citing Sopater (RG 5.79.14f.), but notes that another testimonium (PS 60.13-15 Rabe = RG 2.683 Walz) gives him five (p. xlvii n.1). On p. lxvi he quotes Sopater's prolegomena (RG 5.8.19-21) to the effect that Lollianus fixed the number of issues at seven, and the younger Hermagoras as five, noting that there are inconsistencies in the tradition concerning the number of issues recognised by Lollianus and the younger Hermagoras respectively (p. lxvi n.2). In the absence of cross-references, it is hard to see how readers will be able to assemble these scattered observations together or make coherent sense of them. But the solution is not in doubt: the names of Lollianus and the younger Hermagoras have been transposed in Sopater's prolegomena; Lollianus had five issues (the list at RG 5.79.14f. has lost one member), and the younger Hermagoras seven. Undoing that transposition confirms five issues for Lollianus. The solution was established by Gloeckner in a discussion that Patillon (p. lxvi n.2) cites so vaguely that the reader might infer that Gloeckner had done no more than point out the problem.12
Patillon also says that the younger Hermagoras 'fait des quatre qualités de l'opposition autant d'états de cause' (p. xlvii). That is not consistent with the list of Hermagoras' seven issues at RG 5.79.10-13; confirmation that Hermagoras treated quality (of which antithesis, Patillon's 'opposition', my 'counterposition', is a subdivision) as a single issue with a variety of species can be found at RG 4.223.4-7 -- a passage which Patillon himself cites in another context (p. 101). No supporting reference is given for the statement on p. xlvii, but Patillon presumably had Sopater's observations at RG 5.174.16-24 in mind. There it would be have been more strictly accurate to say that, unlike Lollianus, Hermagoras distinguished the four counterpositions and included them among the species of the single issue, quality; but since Sopater's agenda in that passage was systematic rather than historical, he translated Hermagoras' doctrine into the terminology of the later thirteen-issue system. I do not doubt that Patillon himself understands this; what is puzzling is, again, his failure to integrate the material and to offer the reader help in making sense of the apparent contradictions.
Rabe's edition will remain in front-line service until Patillon's Corpus Rhetoricum is complete. Fortunately, my copy of the 1969 reprint has withstood two decades of intensive use with only slight signs of wear. It also has the merit of staying open on my desk of its own accord. The tightly bound paperback of Patillon's edition is less cooperative and will, I fear, be less resistant to disintegration.
1. Theon (1997); 'Apsines' (2001); the fragments of Longinus, together with a short work by Rufus (2001, with L. Brisson); pseudo-Aristides (2002); the Anonymus Seguerianus (2005).
2. This is revised from the translation in M. Patillon, Hermogène: L'art rhétorique (Paris 1997), which includes all five of the texts transmitted under Hermogenes' name. See also his monograph, La théorie du discours chez Hermogène. Essai sur les structures linguistiques de la rhétorique ancienne (Paris 1988).
3. H. Rabe, Hermogenes (Leipzig 1913), xxvi: 'Interim igitur editionem "minorem" emitto; toto rhetorum graecorum corpore absoluto maiora spectare necesse erit.' The relatively inaccessible edition of G. Kowalski, Hermogenes. De Statibus (Breslau 1947) was deeply flawed.
4. M. Heath, Hermogenes On Issues: strategies of argument in later Greek rhetoric (Oxford 1995).
5. Once ὁ δέ had given way to ἢ, requiring an alternative formulation of the forcible definition, the intrusion of μηδὲ from the previous line was more or less inevitable, and the change of οὐχ to οὐδ' follows as a consequence. These variants, and the omission of ἐστιν, are all characteristic of the branch of the tradition represented by PaPc.
6. 5.3.3 (65.21 Rabe), 5.6.2 (66.7), 8.13.2 (81.17)), 9.7.1 (83.8), 10.18.2 and 3 (87.19f.), 11.5.2 and 6.1 (88.17f.), 11.9.2 (89.11).
7. I commented on this tendency in my reviews of Patillon's Apsines (CR 52 (2002), 11-13) and the Anonymus Seguerianus (BMCR 2005.09.16). So, too, G.A. Kennedy, Gnomon 76 (2004), 306-9.
8. See M. Heath, 'Metalepsis, paragraphe and the scholia to Hermogenes, LICS 2.2 (2003), 1-91 (pp. 16f., 27-9, 31-4), where I argue that the Sopater of RG 4 is a later homonym, who adapted material from the original version of the RG 5 commentary, as well as from other sources. On the Sopater commentary in RG 5 see ibid., pp. 24-7.
9. I discuss the question, and other aspects of the sources for Hermogenes' biography (including the report of his hairy heart), in M. Heath, 'Hermogenes' biographers', Eranos 96 (1998), 44-54.
10. As I noted in 'The family of Minucianus?', ZPE 113 (1996), 66-70 (p. 68), and pointed out in my review of Patillon and Brisson's Longinus, CR 52 (2002), 276-8.
11. The testimonia and fragments are collected in M. Heath, Menander: a rhetor in context (Oxford 2004), 93-131.
12. S. Gloeckner, Quaestiones Rhetoricae (Breslauer Philologische Abhandlungen 8.2, 1901), 52f. The younger Hermagoras (2nd century AD) must not be confused with his more famous homonym (2nd century BC), and we must be cautious in dealing with citations in sources from the third century or later. It has generally been assumed that Porphyry, in RG 4.397.14, was referring to the earlier Hermagoras (= F14a Matthes), and Patillon follows that consensus (p. xlvii, cf. p. 130f.); I think that the reference is more likely to be to the younger Hermagoras. See M. Heath, 'Hermagoras: transmission and attribution', Philologus146 (2002), 287-98, and 'Porphyry's rhetoric', CQ 53 (2003), 141-166 (pp. 150-2).