promoting access to White Rose research papers Universities of Leeds, Sheffield and York http://eprints.whiterose.ac.uk/ White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/2655/ Published paper Heath, M. (1999) Longinus On Sublimity. Proceedings of the Cambridge Philological Society, 45 . pp. 43-74. White Rose Research Online [email protected] Proceedings of the Cambridge Philological Society 45 (1999), 43-74 © Malcolm Heath Longinus On Sublimity MALCOLM HEATH (UNIVERSITY OF LEEDS) Abstract: The traditional attribution of On Sublimity to the third-century critic Cassius Longinus has been rejected by most scholars since the early nineteenth century. The arguments against a third-century date are examined and shown to be unfounded. It is argued that the interest in sublimity and a number of aspects of the treatise’s vocabulary show distinctive points of contact with the evidence for Cassius Longinus, and with authors influenced by him. There is therefore a balance of probability in favour of the traditional attribution. 1. Introduction Until the beginning of the nineteenth century the treatise On Sublimity was universally attributed to the third-century critic, rhetorician and philosopher Cassius Longinus.1 Weiske’s edition, first issued in 1809, marked a turning-point in the trend of scholarly opinion, and Longinus’ claim to authorship is now generally rejected, often summarily.2 A variety of alternative attributions have been canvassed; most commonly the work is assigned to an anonymous author of the first century AD.3 But a minority of scholars have resisted the consensus and defended Longinus’ claim to authorship.4 This paper will argue that they were right to do so. To avoid ambiguity, I shall follow Russell in using the symbol ‘L’ as a non- committal way of designating the author of On Sublimity; by ‘Longinus’ I shall always mean Cassius Longinus. So the question before us is whether L is Longinus. I begin by explaining why manuscript evidence (§2) and stylistic comparison with the fragments of Longinus (§3) fail to resolve the question. I then try to find a place for the composition of the treatise within Longinus’ career 1 RE Longinos (= Aulitzky 1927); PLRE I Longinus (2); FGrH 1091. Brisson and Patillon 1994, 1998, provide a detailed study and a collection of the fragments (cited here in the form F1); but I would dissent from their rejection of the essay on memory (see n.10 below) and the testimonium to the chronographic work (see n.46). On Longinus’ Homeric scholarship (F22-24, F27) see now Dyck 1989, 7f. [Additional note (August 2007): The standard edition of Longinus’ fragments is now M. Patillon and L. Brisson (ed.), Longin. Fragments. Art Rhétorique. Rufus. Art Rhétorique (Paris 2001); my review (Classical Review 52 (2002), 276f.) draws attention to some additional material. There is also an extensive study of the fragments by I. Männlein-Robert, Longin Philologe und Philosoph. Eine Interpetation der Erhaltenen Zeugnisse (Beiträge zur Altertumskunde 143, Munich-Leipzig 2001), which I reviewed in Bryn Mawr Classical Review 2002.03.18. Neither Patillon and Brisson nor Männlein-Robert address the authorship of On Sublimity in any depth. I have set the case for Longinus’ authorship in a larger third-century context in Menader: a rhetor in context (Oxford 2004), 52-89] 2 E.g. Brisson and Patillon 1998, 3104f., devoting four lines of their 172-page study to this question. 3 E.g. Kennedy 1972, 369-72, arguing for an Augustan or early Tiberian date (but see 640 for a reservation; and Kennedy would now opt for a date in the second century: 1997, 34). 4 Notably Marx 1898; Luck 1967; Williams 1978, 17-25. Grube 1965, 340-2 is open-minded. 1 MALCOLM HEATH, LONGINUS ON SUBLIMITY (§4). This leads to a consideration of the final chapter, widely regarded as inconsistent with a third-century date; I shall argue that there is no inconsistency (§5). If so, the way lies open to a reassessment of the case in favour of Longinus’ claim. L’s critical thought and vocabulary prove (§6) to be connected in numerous ways with the fragments of Longinus, and with the works of neoplatonist authors who wrote under his influence.5 Some residual issues are addressed in an Appendix. 2. Manuscript evidence In P (= Par. 2036, the only primary witness) the treatise carries the superscription ‘Dionysius Longinus’. A name of the form ‘Cassius Dionysius Longinus’ (or ‘Cassius Longinus Dionysius’) is unobjectionable in itself. Greeks with Roman citizenship commonly used their Greek personal name as the third of the tria nomina, but there are many examples of a Greek personal name combined with a Latin cognomen; for example, the second-century sophist Dionysius of Miletus was named T. Claudius Flavianus Dionysius.6 But Longinus is nowhere else referred to as ‘Dionysius Longinus’, and the fact that he and Dionysius of Halicarnassus were the two most famous literary critics of antiquity suggests that the superscription may conflate alternative attributions. The disjunction ‘Dionysius or Longinus’ is in fact preserved (or restored) in the list of contents in P. The treatise is certainly not by Dionysius; it is likely that this conjecture was prompted by the reference in 39.1 to the author’s two books On Composition.7 It is possible, therefore, that the attribution to Dionysius arose as a conjectural alternative to an original and authentic attribution to Longinus. But that is not certain: at least one of the alternative attributions must be a false conjecture, but both may be. The manuscript evidence could not, in any event, be decisive. Even if P were unequivocal in attributing the work to Longinus, that attribution might be conjectural and mistaken. There are many examples of rhetorical texts that have at some point been transmitted without an author’s name, and which have acquired incorrect attributions through conjecture.8 The current consensus holds that there is compelling internal evidence against a third-century date for On Sublimity; if so, we should have to reject the attribution to Longinus irrespective of manuscript evidence in his favour. Conversely, the attribution might be conjectural and yet correct: conjectural attributions are sometimes right, as a pertinent anecdote will show. 5 The potential importance of the later Platonists for the problem was stressed by Luck 1967, 99, 112. 6 IEph. 426: see Jones 1980, 373f.; examples could be added indefinitely. 7 A scholion ad loc. notes that Dionysius had written on this topic (perˆ sunqšsewj œgraye DionÚsioj). In the same way, references in the pseudo-Dionysian Art of Rhetoric to the author’s On Imitation (364.24, 383.22 Usener-Radermacher) prompted the attribution to Dionysius of Halicarnassus (see the scholion to 359.2, and p. xxii of the editors’ introduction). 8 To the examples collected in Heath 1998a, 89f. (the paper goes on to question the authorship of the treatise traditionally attributed to Apsines) one might add Quint. 3.5.14, on texts attributed to Hermagoras. 2 MALCOLM HEATH, LONGINUS ON SUBLIMITY 3. Stylistic arguments In 1765 David Ruhnken, reading the rhetorical treatise attributed to Apsines in what was then the only printed edition (the Aldine Rhetores Graeci), realised that a fragment of a work by a different hand had intruded into it, presumably as a result of an incorrectly bound exemplar at an earlier stage of the text’s transmission. This diagnosis was subsequently confirmed by the discovery of a manuscript of the treatise (Par. 1874) from which the intrusive fragment is absent. Ruhnken identified the author of the fragment as Longinus. According to his own account of the discovery this attribution was first suggested by similarities in style to the treatise On Sublimity: ‘J’y reconnus non seulement la marche de Longin, mais plusieurs expressions qui lui sont particulières.’9 Confirmation of the hypothesis came when Ruhnken recognised in the fragment a passage quoted under Longinus’ name in the scholia to Hermogenes; further confirmation was provided by the discovery of an epitome of the same work, again under Longinus’ name.10 Ruhnken’s judgement that there were stylistic resemblances between the fragment of Longinus’ Art of Rhetoric and the treatise On Sublimity has been echoed by scholars who accept Longinus’ authorship of the treatise;11 sceptics have rejected it.12 Partisans on either side of the debate might be tempted to suppose that this divergence of opinion reflects a lack of objectivity on the other;13 but in fact the uncertainty is more deeply rooted. It should, in any case, be clear that this problem needs to be approached with caution. To compare an instance in which common authorship does not seem open to doubt, the two authentic works of Hermogenes illustrates how markedly treatises by a single author can differ in style and manner of presentation. In the present case, we have good reason to expect such a difference: a technical handbook of modest ambitions is unlikely to deploy the same stylistic resources as a more elaborate formal composition.14 It has often been observed that in many passages L adapts his style to the subject-matter in detail, echoing stylistic features discussed in the 9 The discovery was announced in Ruhnken 1765, reprinted in Wyttenbach 1821, 793-5; Wyttenbach’s own narrative (1821, 640f.) is in turn reproduced by Walz (RG 9.xxiii n.19). 10 Fragment: F15b = 179-97 Spengel-Hammer (194.9-18 ~ Longinus ap. RG 5.451.12-452.8; unattributed in John of Sicily, RG 6.119.21-7). Epitome: F15a = 208-12 Spengel-Hammer, re- edited in Gautier 1977, who shows that the epitomator was Michael Psellus. The intrusive material in ‘Apsines’ also includes: (i) an essay on memory (197-206 Spengel-Hammer), the attribution of which to Longinus is contested but in my view probable (on this point I agree with Aulitzky 1927, 1411-3 against Brisson and Patillon 1998, 3042); (ii) some notes on the heads of purpose (telik¦ kef£laia) which there is no reason to associate with Longinus (206-7 Spengel-Hammer). 11 Marx 1898, 195; Luck 1967, 109 (‘Nicht nur sachlich, auch stilistisch ist eine erstaunliche Verwandschaft zu beobachten, die allerdings nicht immer leicht zu beschreiben ist’). 12 Kaibel 1899, 116f.; according to Russell 1964, xxiv-xxv, Longinus’ fragments ‘wholly lack the abundant metaphor and pregnant sententiousness’ of On Sublimity; cf. p. xl, on L’s ‘lavish metaphor and immense richness... a marked fondness for graves sententiae of a Tacitean ring’. 13 A lack of objectivity on Kaibel’s part has indeed been recognised by some who accept his conclusion (Aulitzky 1927, 1406; Walsdorff 1927, 95 n.1) as well as by opponents (Luck 1967, 98). 14 Innes 1994, 48 comments on L’s tendency to treat technical terminology allusively, and on the apology for overly technical treatment in 29.2. 3 MALCOLM HEATH, LONGINUS ON SUBLIMITY text;15 and it seems clear that he has done the same thing on a larger scale in the treatise as a whole, adopting a specially heightened style appropriate to his subject.16 It would be wrong, therefore, to assume that the style of On Sublimity is typical of its author’s manner. Indeed, the treatise itself does not sustain the same stylistic level throughout: consider chapters 40-43, ‘a stretch of writing... in which a certain haste and disorder are apparent.’17 Longinus himself was demonstrably a versatile stylist. The fragment of the Art of Rhetoric (F15b = 179-97 Spengel-Hammer) is not identical in style to the epistolary preface to his treatise On the End (F2 = Porphyry Life of Plotinus 20- 21), the fragment of a philosophical critique of the Stoic doctrine of the soul (F8 = Eusebius Praep. Evang. 15.21), the prolegomena to the commentary on Hephaestion’s treatise on metre (F14a = 81-87 Consbruch), or the lecture on memory (197-206 Spengel-Hammer). If we had his Odaenathus (F6e = Libanius Ep. 1078), presumably an epideictic speech in honour of the Palmyrene king to whose court he moved at some time in the 260s (see §4 below), he would no doubt show us yet another stylistic face. This variety of styles is not surprising: the fragments differ from each other (and from On Sublimity) in genre, and Longinus’ rhetorical training, as well as his critical expertise in questions of style, would have fostered stylistic versatility.18 We have, then, a text written in a style that is unlikely to be typical of its author, and an author who could command a wide range of styles and whose works are represented by relatively small samples of many different kinds of writing—a trivial fraction of his total output. There is surely no way to determine, either positively or negatively, whether this author’s stylistic range could have stretched as far as the style of this text. It is possible that more specific inconsistencies in linguistic practice would enable us to put Longinus’ authorship in doubt. For example, Russell acutely observes that with p£ntej Longinus repeatedly uses ™fexÁj, while L uses ˜xÁj, and comments: ‘this is the sort of detail which makes good evidence for a difference of author.’19 But further investigation suggests otherwise. Aelius Aristides’ practice is strikingly erratic: On the Four uses only p£ntej ™fexÁj (nine times), while Against Plato and the Sacred Discourses use only p£ntej ˜xÁj (twice each); of his other works some use one form, some the other, while the Panathenaicus uses both.20 I cannot fathom the reasons for this variation, but the fact is beyond doubt; so this usage cannot be made an index of authorship. 15 Some examples, and further references, in Innes 1994, 48. 16 Boileau: ‘En traitant des beautez de l’Elocution, il a employé toutes les finesses de l’Elocution. Souvent il fait la figure qu’il enseigne; et en parlant du Sublime, il est lui-mesme tres-sublime’ (preface to Traité du Sublime (1674), in Boileau 1966, 333). Cf. Pope Essay on Criticism 679f., on ‘bold Longinus’, ‘whose own Example strengthens all his Laws, / and is himself that great Sublime he draws.’ The features of style which Russell picks out as characteristic of L (n.12 above) are precisely those which an author aiming for this effect might cultivate. 17 Russell 1964, 185. 18 On stylistic versatility in the second sophistic see Pernot 1993, 1.336-8. 19 Russell 1964, xxv n.1. Note that in this paragraph I am using ‘p£ntej’ as a shorthand for ‘p£ntej and cognates’. 20 Consider again Hermogenes, who uses only p£ntej ˜xÁj (twice) in On Issues, and only p£ntej ™fexÁj (once) in On Types of Style. 4 MALCOLM HEATH, LONGINUS ON SUBLIMITY Quantifiable aspects of style, such as hiatus, rhythm, sentence-length and structure, are also unlikely to prove helpful: the sample-size is too small, and we know that these features were deliberately varied to achieve different stylistic effects.21 I do not believe, therefore, that in this instance the question of authorship can be resolved on stylistic grounds. That means, of course, that Ruhnken’s conjecture about the authorship of the fragment on rhetoric could not safely have been accepted without other evidence to confirm it. But other evidence was forthcoming. So Ruhnken’s sense of style has at least this to be said for it: it produced a hypothesis capable of independent confirmation, that was in fact confirmed. The coincidence of what would (if L is not Longinus) be a correct attribution based on comparison with a misattributed text should perhaps have worried scholars more than it has done. There is, at least, a prima facie case for re- opening the question. 4. The treatise and Longinus’ career Longinus was probably born between AD 200 and 213.22 As an adolescent he travelled widely with his parents, and studied philosophy with a number of teachers, spending most time in Alexandria with the Platonists Ammonius and Origen (F2 = Porphyry Life of Plotinus 20).23 He subsequently settled in Athens, where his maternal uncle, the otherwise unknown Fronto of Emesa, taught rhetoric; Longinus was Fronto’s heir (F1b = Suda F735). In Athens he taught literature, rhetoric and philosophy. He moved to Palmyra in the mid- or late- 260s,24 and was an adviser to Zenobia. When Aurelian captured Palmyra in 273 Longinus was executed on suspicion of complicity in her revolt (F6b = Zosimus 1.56.2f.). This last stage of his career has tended to dominate perceptions of Longinus. Edward Gibbon wrote in his journal for 11th September 1762:25 21 See, in particular, their role in Hermogenes’ analyses of the types of style. For rhythm as a generic variable see Berry 1996. 22 These approximate termini are established by Brisson and Patillon 1994, 5219f. The date usually given (c. 213) is based solely on the assumption (which attains the status of a fact in Longinus’ entry in PLRE I) that he was about twenty years older than his pupil Porphyry. 23 On the confusing prosopography see Edwards 1993. Ammonius, teacher of Plotinus and the neoplatonist Origen, must be distinguished from Ammonius, teacher of the Christian Origen. The former is the Platonist with whom Longinus also studied. Edwards suggests that the latter is the Peripatetic mentioned by Longinus (ibid.) as a distinguished scholar and polymath. But his only literary remains, according to Longinus, were ‘poems and epideictic speeches’; contrast the many works attributable to Origen’s teacher according to Eusebius (HE 6.19.10). Nor is there any reason to believe that the Peripatetic Ammonius (known also to Philostratus: VS 618) was based in Alexandria. (The Ammonius mentioned by L (13.3) has, of course, nothing to do with any of these: see (e.g.) Russell ad loc.) 24 If his Odaenethus (F6e = Lib. Ep. 1078) was an epideictic speech in honour of the Palmyrene king, Longinus must have been in Palmyra before (or, if it was a funeral speech, not long after) Odaenethus’ death in 267. 25 Low 1929, 139. 5 MALCOLM HEATH, LONGINUS ON SUBLIMITY When I reflect on the age in which Longinus lived, an age which produced scarce any other writer worthy the attention of posterity, when real learning was almost extinct, Philosophy sunk down to the quibbles of Grammarians and the tricks of mountebanks, and the Empire desolated by every Calamity, I am amazed that at such a period, in the heart of Syria, and at the Court of an Eastern Monarch, Longinus could produce a work worthy of the best and freest days of Athens. Weiske, too, when he launched the case against Longinus’ authorship referred specifically to the age of Aurelian (AD 270-275).26 But there is no reason to suppose that the treatise (if Longinus wrote it) was a product of his last years. It is more likely to have been written when he was teaching in Athens. The author has been reading Caecilius with Terentianus (1.1), a young man (15.1); the obvious inference is that Terentianus was a pupil.27 The addressee of Longinus’ On the End was a Roman named Marcellus (F2 = Porphyry Life of Plotinus 20), most probably a pupil.28 One might compare book 3 of the pseudo-Hermogenean On Invention,29 addressed to a former pupil named Marcus Julius (126.2-4). Needless to say, the address to a Roman pupil does not imply that the author himself taught in Rome; Romans continued to go to Athens to study in the imperial period (one thinks, for example, of Apuleius and Aulus Gellius in the second century). Longinus, by all ancient accounts, was a major figure in the intellectual landscape of his time. Porphyry, the most distinguished of his pupils, describes him as the greatest critic of the age (F5 = Life of Plotinus 20: toà kaq' ¹m©j kritikwt£tou genomšnou). For Eunapius he was ‘a living library, and a research institute on legs’ (biblioq»kh tij Ãn œmyucoj kaˆ peripatoàn mouse‹on), a critic on a par with Dionysius of Halicarnassus, and by far the most distinguished man among his contemporaries, to whose critical judgement everyone deferred (F3a = Eunapius Lives of the Sophists 4.1.1-6 (6.9-7.7 Giangrande)). A glimpse of Longinus’ intellectual milieu is provided by a fragment of Porphyry, describing a dinner given by Longinus on Plato’s birthday (F4 = Eusebius Praep. Evang. 10.3 (Porphyry 408F Smith)); since Porphyry is one of those present, the dramatic date must be before 263, when he left Athens. At the top table, with Longinus himself, are the rhetoricians Nicagoras and Maior, the grammarian Apollonius, the geometer Demetrius and the philosophers Prosenes (a 26 See n.58 below. 27 The name Terentianus is secured by 1.4 etc., but at 1.1 P reads PostoÚmie Flwrentianš. Manutius’ PostoÚmie Terentian, printed by most editors, does not account satisfactorily for the transmitted reading. Marx 1898, 182-4 (cf. also Luck 1967, 102f.) advocates Schurzfleisch’s PostoÚmie Fl. Terentian, and suggests a connection with the Flavii Postumii attested in the third century (T. Flavius Postumius Titianus = PLRE I Titianus (9); T. Flavius Postumius Varus = PLRE I Varus (2)). 28 He has been identified plausibly with the senator Marcellus Or(r)ontius, who also studied with Plotinus (Life of Plotinus 7): cf. RE Suppl. 15, Marcellus (9a); PLRE I Orontius; Alföldi 1967, 256; Brisson et al. 1982, 96f. 29 By Longinus’ older contemporary Apsines (who also taught in Athens), if the conjecture in Heath 1998a is correct. 6 MALCOLM HEATH, LONGINUS ON SUBLIMITY Peripatetic) and Callietes (a Stoic).30 An argument between two other guests about the relative merits of the historians Theopompus and Ephorus prompts a wide- ranging discussion of literary plagiarism. The discussion touches at one point on Plato (10.3.24), to whom Prosenes is reluctant to apply that term; one may recall L’s eagerness (13.4) to distinguish Plato’s debt to Homer from plagiarism (klop»). In the course of this discussion Apollonius cites a judgement of Caecilius on Menander (10.3.13). So Caecilius’ critical opinions were available to people in Longinus’ intellectual circle in Athens, and of interest to them. This does not prove that Caecilius’ writings were available (his opinions could have been known at second-hand), but there is no reason to doubt that they were. Caecilius was one of the main sources for Tiberius On Figures.31 Moreover, a collection of excerpts from Longinus (F16 = 213-6 Spengel-Hammer) contains observations on the classical orators that are paralleled in Photius, whose fuller version in one instance names Caecilius; the prima facie implication is that Photius derived material (probably indirectly) from Longinus, and that Longinus quoted and commented on Caecilius’ views.32 Since we know that Longinus was interested in sublimity (the evidence is presented in §6 below) it is likely that he would have read Caecilius’ treatment of the subject. L had written on sublimity before he wrote the extant treatise (at 9.2 he quotes his own earlier formulation); this antecedent interest is no doubt the reason why he chose to read Caecilius on sublimity with his pupil Terentianus. If Longinus is L, these data come together very neatly.33 Another work of Caecilius that was no doubt of interest to L was his comparison of Cicero and Demosthenes (fr. 153 Ofenloch = Plutarch Dem. 3). We do not know whether L had direct knowledge of Cicero; his comparison of the two orators (12.4f.) may be wholly derivative. So the identification of L with Longinus may be maintained whatever the likelihood of a third-century Greek scholar having first-hand acquaintance with Cicero. First-hand acquaintance cannot, in fact, be excluded. Gellius (10.9.7) attests to Greeks acquainted with (although not enthusiastic about) Latin literature in the second century. Longinus’ near- contemporary Gregory Thaumaturgus studied Latin in his youth, ‘not with a view 30 Demetrius and Apollonius were probably also among Porphyry’s teachers: for Apollonius see Porphyry QH 1.111.9f. Sodano = 14T Smith; for Demetrius (RE Demetrios (118)) see Proclus In Remp. 2.23.14f. = 13T Smith. Nicagoras was probably related to the Minucianus who was a rival of Hermogenes, and on whose Art of Rhetoric Porphyry wrote a commentary: see Heath 1996. Maior wrote a work on stasis-theory, of which a few fragments are preserved. Prosenes and Callietes are not otherwise attested; nor are Caystrius and Maximus, the other guests mentioned. 31 Tiberius (probably the philosopher and sophist of Suda T550) also cites Apsines, and is therefore T not earlier than the third century. Cf. Solmsen 1936; Ballaira 1968. 32 The ascription of the excerpts to Longinus has been disputed; I concur with Aulitzky 1927, 1411, and Brisson and Patillon 1998, 3078-80. I discuss the possible connection between Longinus and Photius more fully in Heath 1998; Treadgold’s suggestion (1980, 50 n.53) that Photius drew on Proclus’ Chrestomathy in his discussions of the ten orators would make good sense of the presence of Longinian material, in view of the interest shown in Longinus by Syrianus and his pupils Proclus, Hermias and Lachares (§6 below). 33 The idea that L’s reply must have been written close in time to Caecilius cannot be taken seriously: Plutarch wrote against Colotes (as Marx 1898, 194 observed; cf. Russell 1989, 309); Aelius Aristides wrote replies to Plato. It may also be mentioned in passing that the reference to Theodorus (presumably, of Gadara) in 3.5 does not imply that L was his pupil: Grube 1959, 356- 65. 7 MALCOLM HEATH, LONGINUS ON SUBLIMITY to perfect fluency, but so that I should not be wholly unacquainted with this language also’; and he was doing so before he thought of taking up Roman law, since it was his Latin tutor who suggested that he do so (Panegyric on Origen 56- 9, PG 10.1065). The allusive etymology in Herodian 1.12.2 assumes a readership acquainted with Latin. Eusebius of Caesarea was able to translate documents from Latin. And there is evidence for bilingual texts of Cicero in the fourth century, at least.34 But even if Longinus had no direct acquaintance with Cicero’s works, there is no doubt that he would have been aware of his literary significance. In the third century Cassius Dio shows knowledge of Cicero’s speeches;35 he perhaps is a special case, but an allusion in Philostratus’ Life of Apollonius (7.11f.) discloses, and assumes on the part of the reader, an awareness of Cicero as an author of philosophical dialogues.36 In the third or fourth century Aristides Quintilianus refers (2.6) to Cicero’s Republic,37 and the following discussion shows knowledge of Pro Roscio Comoedo and the Tusculan Disputations (4.1.3). In (probably) the fourth century Sopater refers to Cicero as evidence that rhetoric had not entirely disappeared in the low period between the establishment of Macedonian hegemony and the renaissance of the second century AD (RG 5.8.11-18 Walz); this is part of an outline history of rhetoric that may derive indirectly from the commentary on Minucianus by Longinus’ pupil Porphyry.38 Another text beyond the usual range of reference of Greek critics but mentioned by L (9.9) is Genesis. Longinus is named as an admirer of Moses by John of Sicily (RG 6.211.12-15), but this testimonium must be treated with reserve, since it may simply reflect John’s belief that Longinus wrote On Sublimity. Although John does preserve information about Longinus’ Philological Discourses elsewhere,39 in one other passage there is reason to believe that a reference to the Philological Discourses in John’s source has been conflated with his own recollections of On Sublimity (see §6.3 below). Even so, L’s reference to Moses fits easily into Longinus’ intellectual background. At least three second- century pagan authors display familiarity with the Genesis cosmogony: Celsus regards it as an absurd myth, and subjects it to polemical attack; Galen evaluates it as a philosophical position, superior to Epicureanism though less complete than Platonist cosmology; Numenius (with whose work Longinus was certainly familiar) cites it with respect.40 34 Fisher 1982, 183-5; Rochette 1997, 188-98 on ‘translations’ of Virgil and Cicero (also 263-9: Greeks and Roman literature; 279-86: Greeks and Cicero). 35 Gowing 1992, 237f.; Gowing 1998. 36 Gowing 1998, 384-7. 37 He may have derived the information from Suetonius’ work (in Greek) on Cicero’s Republic (Suda T895). T 38 Rabe 1931, xi-xiv. On Porphyry’s rhetorical writings see Heath 2002, 2003. 39 RG 6.93.7-94.2 (= F19), 95.1f. (= F21b), 225.9-29 (= F21e); and see n.10 for John’s (unattributed) quotation from Longinus’ Art of Rhetoric. John’s references to Longinus are discussed in Mazzucchi 1990, this one at 187-9. 40 Celsus: Origen Against Celsus 6.49-51, 60-61. Galen: De Usu Partium 3.905 Kühn; cf. Walzer 1949. Numenius: fr. 30 des Places (quoted by Porphyry On the Cave of the Nymphs 10); see also frr. 1, 9-10, 56. For Longinus’ familiarity with Numenius’ works see the preface to On the End, quoted in Porphyry’s Life of Plotinus 20 = F2. Other Greek authors were aware of Moses: see 8 MALCOLM HEATH, LONGINUS ON SUBLIMITY A related question arises from the parallels to Philo of Alexandria that have often been noted in On Sublimity. It has been argued that these parallels reflect knowledge of Philo on L’s part; the case is not conclusive,41 but poses the question what bearing this would have on the identification of L with Longinus. We cannot expect to find confirmation that Longinus was familiar with Philo in the fragments;42 but we may ask whether it is plausible in principle. It is certainly true that pagan Platonists were not normally familiar with Philo;43 but that is true of pagans in the first century as well as the third, so familiarity with Philo would pose no greater problem for the identification of L with Longinus than for any other hypothesis. The abnormality would at least be intelligible in Longinus’ case: the man Eunapius described as ‘a living library, and a research institute on legs’ must have been exceptionally well-read.44 The Christian theologian Origen did not regard it as unthinkable that interested pagans should read and admire Philo. When he suggests that Celsus alludes to Philo and other Jewish allegorists without having read them (Against Celsus 4.51) we should not necessarily believe him; the accusation that an opponent has not read the books he criticises is an obvious polemical device (compare the malicious insinuation at 4.42 that Celsus had not read Genesis), and Celsus had certainly read some non-canonical literature.45 More significant is the fact that Origen treats Celsus’ ignorance of Philo as a matter for satirical conjecture, rather than as something self-evident. Moreover, he suggests that Philo’s style and content would command the respect of philosophers; similarly in 6.21 he commends Philo’s On Dreams to inquirers. What would have motivated such inquirers? If Celsus did read Philo, it was of course to gather material for his polemic; similarly with the study of gnostic literature by Plotinus’ pupils Amelius Gentilianus and Porphyry, and of the Bible Gager 1972; Stern 1976-84, who discusses inter alios L (1.361-5), Numenius (2.206-16) and Galen (2.306-28); Rinaldi 1989. 41 Russell 1964, xxix-xxx. Norden 1954 argues (on the gratuitous assumption that the last chapter of On Sublimity records an actual discussion) that L met Philo in Rome; Kaibel 1899, 130 n.2 denies dependency; Runia 1986, 305f. is non-committal. 42 Plotinus, who offers far more material for comparison, has elicited radically different opinions. Responding to the similarities noted by Armstrong 1940, 70-74, 107-8, Rist 1967, 99-101 is sceptical ‘on a priori grounds’; contrast Gatti 1996, 12, to whom Philo’s influence on Plotinus seems ‘undeniable’. For a judicious recent discussion see Meijer 1992, 326-8. Armstrong 1960, 393f. notes one possible parallel between Philo and Longinus, but concludes that it is superficial and probably fortuitous. 43 In the case of Plotinus, my own view (for what it is worth) is that there is no evidence of familiarity; and I agree with Edwards 1990, who warns against over-estimating the extent and depth of Numenius’ knowledge of Jewish thought. [M. Burnyeat, ‘Platonism in the Bible: Numenius of Apamea on Exodus and eternity’, in R. Salles (ed.), Metaphysics, Soul, and Ethics in Ancient Thought (Oxford 2005), 143-69 takes a more positive view of Numenius’ knowledge of the biblical texts.] 44 I note that the pagan grammarian Helladius (early fourth century) cited Philo (Photius cod. 279, 529b27-29); see Gager 1972, 129-31. 45 See Against Celsus 4.52, on Jason and Papiscus. Andresen 1955 makes a strong case for Celsus’ familiarity with the works of Justin; on Celsus and Philo see Stein 1932/3. Celsus’ polemical purposes gave him as much motive to conceal the existence of intellectually respectable Jewish literature as Origen had to deny that Celsus had read it. I note in passing that Amelius Gentilianus referred favourably to the beginning of John’s gospel (Eusebius PE 11.19); see Dörrie 1972. 9