DRN 4.1171–91: THE ERRORS OF COMEDY

At 4.1171–91 we find a satirical Epicurean take on the motif of the exclusus amator. While a male lover abases himself before her doors, the woman hides within, fumigating herself (1177–84).Footnote ⁵ In the lines that follow, the alleged tendency of women to hide themselves away in order to prevent their male lovers from recognizing their true natures is introduced with an explicitly theatrical metaphor:

The term postscaenia at 1186 is hapax legomenon.Footnote ⁶ It has been suggested that it is a Lucretian coinage;Footnote ⁷ alternatively, it could provide evidence for a technical term from theatre architecture not attested elsewhere in Latin.Footnote ⁸ The phrase of which the term is a part (uitae postscaenia, ‘things behind the scenes of life’) is a metaphor for those areas of female life to which male lovers are not granted access. In the light of its strong theatrical connotations, however, we may also read it as a generic marker, pointing to the theatrical significance of the stock scene that has just been described.Footnote ⁹ The relevance of the Roman theatre to Lucretius' depiction of the exclusus amator is confirmed by a number of clear points of contact between Lucretius' scene and the paraclausithyra of Plautus and Terence:Footnote ¹⁰ the use of the verb excludo (1177);Footnote ¹¹ the weeping lover (1177);Footnote ¹² the personification of the door/doorposts (1178);Footnote ¹³ the kissing of the door (1179);Footnote ¹⁴ the anointing of the door (1179);Footnote ¹⁵ the characterization of the beloved as Venus (1185).Footnote ¹⁶ The theatrical metaphor uitae postscaenia serves to activate the reader's recognition of these structural similarities, and in this way the reader is cast as a spectator of a comic scene, observing in Lucretius' exclusus amator a quasi-theatrical representation of a man in love.Footnote ¹⁷

The exclusus amator of 4.1171–91 serves as an object lesson for the student of Epicurean rationalism. Erotic love, for Lucretius (see 4.1149–54), involves a dissonance between the lover's perceptions (which are, of course, true) and his mistaken interpretation of those perceptions—a dissonance which gives rise to false beliefs concerning the nature and attributes of the beloved. This model of the confusion resulting from erotic love is in full alignment with the Epicurean account of cognitive error,Footnote ¹⁸ according to which error is said to lie not in perceptions themselves but rather in the opinions that are added to them.Footnote ¹⁹ The error of a man in love with a woman who really is beautiful, such as the exclusus amator of 4.1171–91, consists in his failure to realize that she engages in the same deceptive practices as all other women (4.1171–6). If the true nature of her behind-the-scenes activity were revealed, he would at once recognize his error and depart in peace, seeing that he had attributed to her more than was warranted: tribuisse quod illi | plus uideat quam mortali concedere par est (4.1183–4). In Epicurean terms, he has formed beliefs concerning the nature of his beloved that were not warranted by his limited perceptual experience of her.

This account of the exclusus amator's error neatly demonstrates a central principle of Epicurean rationalism: in cases where we do not have sufficient evidence to form an accurate opinion, we should not commit to a view until better-quality evidence becomes available, at which point our candidate opinions can be confirmed or denied.Footnote ²⁰ This is the doctrine of τὸ προσμένον, ‘the opinion awaiting confirmation’.Footnote ²¹ By depicting as an exclusus amator a deluded lover who has wrongly committed himself to opinions founded on incomplete evidence, Lucretius invites his readers to reconfigure, along Epicurean lines, their understanding of a stock scene of comic theatre: the suffering of the exclusus amator is rooted not in his being denied access to his beloved but rather in his being denied access to good evidence about her true nature—evidence on the basis of which he could have avoided forming false beliefs. As such, Lucretius at 4.1171–91 asks his readers both to recall and to reinterpret the paraclausithyra they have encountered on the Roman stage: not only is the reader's knowledge of theatrical intertexts called upon in order to help dramatize Lucretius' account of erotic love, but those intertexts are subsequently reinterpreted according to an Epicurean rationalist model of erotic suffering and error.

THE PROBLEM OF TRAGEDY

The genre of tragedy represents a more difficult domain of cultural experience for the rationalist to negotiate. Although tragic plots are by no means limited to the world of myth, it is, none the less, the case that a large number of texts belonging to that genre include mythical stories, the contents of which are antithetical to Epicurean rationalism about how the world works. At the same time, the canonical status of many such texts means that Lucretius' readers are likely to be familiar with their content, and so must be taught to interpret them appropriately. The result is a peculiarly ambivalent allusive strategy (sometimes called ‘demythologization’),Footnote ²² according to which Lucretius both alludes to such texts—together with the mythical stories they contain—and rejects their literal truth, either by baldly asserting a rationalized alternative or by transforming their contents into allegory.Footnote ²³ In the rest of this paper I discuss, in turn, several tragic allusions in DRN, in order to make clear the variety of different responses to tragedy that were available to the Epicurean rationalist. While I will not question the central importance of demythologization to Lucretian tragic intertextuality, I will seek to demonstrate that it does not follow from the presence of irrational mythic elements in a text (or group of texts) that the same text (or group of texts) as a whole should be thought antithetical to Lucretian rationalism. As a result, any model of the relationship between Lucretian rationalism and Lucretian intertextuality must be sufficiently fine-grained as to distinguish between rational and irrational elements within a single text, or within a single genre.

In a recent paper on tragic intertextuality in DRN, Robert Cowan reads Lucretian allusions to tragic texts chiefly at the level of genre.Footnote ²⁴ After identifying Lucretius' phrase odere timentque (3.73) as an ‘anti-allusion’ to the phrase oderint dum metuant from Accius' Atreus (Trag. 203–4 R³ = fr. 10 Dangel), Cowan draws a conclusion about Lucretius' attitude to the relevance of that play's genre to the interpretation of 3.73: ‘By asserting the “reality” of ueneficium in contemporary Rome, Lucretius implicitly denies that of Thyestean cannibalism and in so doing rejects the explanatory power and relevance of the tragic genre.’ (116). Cowan likewise identifies Lucretius' depiction of the sacrifice of Iphianassa as an ‘appropriation’ not merely of a particular play, or group of plays, but of tragedy as a genre (117). While I suspect that the idea of ‘anti-allusion’ at 3.73 is itself problematic,Footnote ²⁵ in this paper I am chiefly concerned with questioning the notion that a Lucretian rejection of a text, or of part of a text, may entail a concomitant rejection of the genre of which that text is a member. Instead, I will argue, pace Cowan, that tragic intertextuality in Lucretius tends to resist analysis at the level of genre. This resistance follows directly from Lucretius' primary didactic aim.

In what follows, I begin with a simple example of Lucretian demythologization of a dramatic text (Accius' Clutemestra), showing how the process of demythologization can emerge from the structure of Lucretius' arguments. This is a case in which the contents of a given dramatic text are straightforwardly antithetical to the rationalist stance of DRN and, as such, may be rejected outright. The following examples (Pacuvius' Chryses, Aeschylus' Agamemnon, Euripides' Iphigenia Aulidensis, Ennius' Andromache) are more complex. In these cases I discuss texts or groups of texts in which both sides of a rational/irrational dichotomy are explored. Lucretian engagement with such texts must go further than simple demythologization. As I will argue, these Lucretian allusions encourage readers to adopt a kind of selective ambivalence to the contents of tragedies—a critical attitude which involves the acceptance of those elements of tragic texts which can be aligned with Epicurean rationalism together with a rejection of any irrational elements also encountered there.

DRN 6.387–95: ACCIUS' CLVTEMESTRA

A straightforward example of Lucretian demythologization of a tragic text is found at 6.387–95:

This is the first of a series of arguments (6.379–422) against the notion that the gods control thunderbolts and use them in order to intervene punitively in mortal affairs.Footnote ²⁶ The argument is modus tollens, with the major premiss and conclusion both implicit in the rhetorical question (this is typically Lucretian): if the gods sent thunderbolts at will, they would send them only upon the guilty; they do not send them only upon the guilty; therefore, the gods do not send thunderbolts at will. The negated lines 391–2 are closely imitated (minus the negation) by Virgil at Aeneid 1.44, whereon Servius Auctus preserves a line of Accius' Clutemestra (Trag. 34–5 R³ = fr. 3 Dangel):

The unusual anatomical specificity of pectore at 6.392, coupled with the close proximity of flammas at 391 and the mention of both Jupiter and alii diui at 387 (surely hinting at Athena—by far the most conspicuous example of ‘another god’ who uses Jupiter's thunderbolt),Footnote ²⁷ confirms this passage of Accius' Clutemestra as a ‘latent’ tragic intertext for DRN 6.387–95.Footnote ²⁸ The Accian line describes Athena's punishment of Locrian Ajax for the rape of Cassandra. Unlike the Homeric tradition, where Ajax is drowned by Poseidon (Od. 4.499–511), Accius follows the story encountered in Athenian tragedy (Eur. Tro. 77–86), where Athena obtains the use of her father's thunderbolt (cf. Aesch. Eum. 827–8) and uses it to destroy Ajax herself.Footnote ²⁹

The demythologization of the Accian account emerges from the structure of Lucretius' argument: the myth of the destruction of Locrian Ajax by Athena asserts that the gods use the thunderbolt to punish the guilty (the consequent of the major premiss). By simultaneously invoking and negating the content of Accius' text (391–2: non faciunt ut …), Lucretius asserts the minor premiss (that is, the guilty are not in fact singled out) and in so doing denies the consequent of the major premiss. The (tacit) conclusion follows. So, the denial of the consequent involves the rejection of part of the mythical tradition as it is recorded in Roman tragedy. The didactic potential of an allusion to this play is clear: a lavish production of Clutemestra was staged for the opening of the theatre of Pompey in 55 b.c. (Cic. Fam. 7.1.2), making it a particularly salient cultural touchstone for both poets and readers in the mid to late fifties b.c. Given that the theatre is an important element of the constructed reader's everyday sensory experience in DRN, it makes sense that Lucretius should call on familiar dramatic texts in order to reconfigure the readers' knowledge of natural phenomena. At the same time, the reader is warned against accepting uncritically the mythical content of tragic texts. This passage, then, in implicitly rejecting the Accian account of the death of Locrian Ajax, constitutes a classic case of Lucretian demythologization.

DRN 1.84–100: AESCHYLUS' AGAMEMNON, EURIPIDES' IPHIGENIA AULIDENSIS, ENNIUS' ANDROMACHA

While Lucretius has a clear rationalist motivation implicitly to reject as false the story of Athena's revenge on Locrian Ajax, it would be wrong to say that he rejects the literal truth of all myths. Some mythical stories are presented in DRN as though they contain historical facts: the Trojan War (1.464–77; 5.326) and the Theban War (5.326) are clear examples.Footnote ³⁰ It does not follow from the supposed historicity of these myths, however, that texts featuring them are fully immune to Lucretian demythologization. This is because even those myths whose ‘basic historicity’Footnote ³¹ is never in question tend to feature a high degree of supernatural content in their canonical literary retellings. As a result, demythologization will often be required even of those texts that narrate stories that are thought basically true. A particularly complex example of such demythologization is provided by the sacrifice of Iphigenia at Aulis, cited by Lucretius (1.84–100, using the alternative name ‘Iphianassa’), as an infamous historical example (1.80–3, 101) of the evils caused by religio:

The contents of the mythic story of the sacrifice of Iphigenia are highly unstable. While Aeschylus, Sophocles and Euripides (in Electra) all state that she died at Aulis, an alternative tradition—that she was rescued by Artemis at the last minute and replaced by a proxy—is attested as early as the Hesiodic Catalogue of Women (fr. 23[a] M-W) and was likely present in the Cypria.Footnote ³² A version of this tradition according to which a deer was substituted for the girl is used in Euripides' Iphigenia in Tauris as an explanation of how Iphigenia escaped from Aulis (28–9; 783–6).Footnote ³³ Within the dramatic tradition, therefore, there existed (in Epicurean terms) both rational and irrational variants of the same myth: some variants included divine intervention on the part of Artemis, which naturally could never be integrated into an Epicurean world view; some variants included no such intervention, and could easily have formed part of a rationalized Epicurean account of mythic history. Roman readers would have been aware of both traditions: the tradition that Iphigenia herself was sacrificed is well represented in both literary and visual art,Footnote ³⁴ while Pacuvius' Chryses—a staple of the Roman theatrical repertoire in the mid first century b.c.,Footnote ³⁵ with which Lucretius directly engages in Book 5 (see below)—relied for its plot on the idea that Iphigenia did not die at Aulis, and seems even to have included a recognition scene between Iphigenia and Orestes.Footnote ³⁶ The Lucretian account at 1.84–100, which foregrounds the death of Iphigenia/Iphianassa as a historical fact at 84–5, asserts the truth of the rational variant and so implicitly rejects the truth of the ‘irrational’ rescue-story.Footnote ³⁷ That readers of DRN should be in no doubt that Iphigenia died at Aulis is, of course, of central importance to Lucretius' argument: this is his first, and most strident, condemnation of the evils wrought by superstition. In asserting the rational variant of the myth (and implicitly rejecting the irrational variant), he teaches his readers how correctly to respond to mythic variation. We should not be surprised, therefore, to detect in Lucretius' version of the story affinities with the account offered in Aeschylus' Agamemnon (a play featuring the rational variant of the myth):Footnote ³⁸ in both the Lucretian (95–6) and the Aeschylean (234–5) accounts, the unwilling girl has to be carried to the altar (contrast the willingness of Iphigenia in Euripides' IA and Ennius' Iphigenia);Footnote ³⁹ in both accounts she goes to the altar in silence, in Aeschylus (235–8) because she is gagged, in Lucretius (92) because she is muta metu; Lucretius' phrase prima uirorum (86) echoes Aeschylus' ἄνθος Ἀργείων (197).Footnote ⁴⁰ These affinities with Aeschylus' Agamemnon support Lucretius' assertion at 84–5 of the rational variant of the myth. This passage, therefore, speaks against any notion that the tragic genre is, generally speaking, at odds with Lucretius' rationalizing project. Far from demythologizing the contents of Aeschylus' play, here Lucretius enlists the Agamemnon as an aid in the implicit rejection of alternative variants of the myth.

Agamemnon, however, is not the only tragic version of the Iphigenia myth to which allusive reference is made in this passage. DRN 1.94 was recognized by Lambinus as a Lucretian rendering of Euripides, IA 1220: πρώτη σ’ ἐκάλεσα πατέρα καὶ σὺ παῖδ’ ἐμέ. To this we may add 1.100, where exitus ut classi felix faustusque daretur closely recalls Achilles' words as reported at IA 1575—καὶ δὸς γενέσθαι πλοῦν νεῶν ἀπήμονα—with the hendyadic formula felix faustusque rendering the tragic word ἀπήμονα, exitus … classi standing in neatly for πλοῦν νεῶν, and the final clause with daretur replacing Achilles' imperative δὸς γενέσθαι.Footnote ⁴¹

Determining which version of the myth Lucretius would have found at the end of Euripides' IA is not straightforward: while the text of that play, as it has come down to us, does feature the rescue-motif at its conclusion, this is not the text that would have been known to Lucretius. Our text of IA contains a large amount of interpolation, occurring on at least two, quite different levels. First, a number of lines, including the whole first half of the exodos (1532–77, classified by Diggle as vix Euripidei),Footnote ⁴² have been attributed to a fourth-century b.c. ‘revival’ of the play.Footnote ⁴³ Second, the remainder of the exodos (1578–629, classified by Diggle as non Euripidei) has been dated much later, to between the fourth and seventh centuries a.d. Footnote ⁴⁴ While it is impossible for Lucretius to have encountered the second, much later level of interpolation (1578–629), it is quite plausible that the first level of interpolation—dated to the fourth century b.c. and surviving in all extant manuscripts—would have been present in texts available to Lucretius. In fact, Lucretius' own account of the sacrifice strongly suggests that he was familiar with a text of the play that included the fourth-century b.c. interpolations: as we have seen, exitus ut classi felix faustusque daretur (1.100) closely recalls IA 1575, which is found in the first half of the exodos.

While all explicit references to the miraculous substitution of the deer in the extant text of IA are found in the much later interpolations of the second half of the exodos (which Lucretius could not have known), the words spoken by the messenger to Clytemnestra at 1537–8 (located in the first half of the exodos) make it quite clear that, in the fourth-century b.c. version probably known to Lucretius, the messenger did indeed go on to narrate the miraculous substitution: in response to Clytemnestra asking whether he has come to tell her news of disaster, the messenger responds with the adversative particle combination μὲν οὖν, saying that, on the contrary, he has come to deliver ‘strange and wondrous news' about her daughter.Footnote ⁴⁵ We can say with some confidence, therefore, (a) that Lucretius was familiar with a version of the Euripidean IA which included the interpolations traditionally dated to the fourth century b.c.,Footnote ⁴⁶ and (b) that said edition featured the irrational variant of the myth, i.e. the miraculous substitution of the girl at the end of the play.Footnote ⁴⁷

How, then, given Lucretius' commitment to teach his readers to think rationally, should we explain the presence in 1.84–100 of allusions to a text (Euripides' IA) featuring the irrational variant of the Iphigenia myth? As has already been stated, there can be no room for doubt in Lucretius' account: convincing his readers that the gods do not intervene in human affairs is crucial to enabling their future happiness. We would be wrong, therefore, to identify in this passage any tension between the rational and the irrational variants—Lucretius, after all, states quite clearly that the girl was killed at Aulis. Rather, we should understand Lucretius here as enabling his readers, with the help of multiple intertexts, to enact a comparison between variant accounts of a single myth, while at the same time explicitly stating which elements should be considered true and implicitly rejecting those elements which are to be considered false. In didactic terms we may say that the purpose of the Euripidean intertexts at 1.84–100 is one of inoculation: by incorporating into his own rationalist account lines taken from a text known to feature the irrational variant, Lucretius demonstrates to his readers how to deal with the interpretative problems posed by mythic variation. The Euripidean lines are removed from their potentially problematic source and inserted into a univocal, fully rationalized discursive environment—DRN—where no room is left for the possibility of divine intervention. The irrational ‘Euripidean’ account is thus corrected in the mind of the reader.

Finally, the lines in which Iphigenia's/Iphianassa's historical death is asserted (84–5 aram … | turparunt sanguine) are not themselves free of tragic allusion. They closely recallFootnote ⁴⁸ Ennius 94 Jocelyn, from a speech in the tragedy Andromacha, spoken by the title character and recorded by Cicero at Tusculanae disputationes 3.45:

That DRN 1.84–5 would recall these lines to an ancient reader is confirmed by Virgil, in whose account of the death of Priam (Aen. 2.499–505) elements of the Ennian and Lucretian passages appear to be combined.Footnote ⁴⁹ Crucially for Lucretius' purposes, the Ennian lines are taken from the same mythic cycle as the story of Iphigenia/Iphianassa, but differ in so far as the death they describe (Priam's) may be called uncontroversial: there exist no mythic variants in which Priam survives the sack of Troy by divine intervention. Furthermore, Iphigenia's sacrifice and Priam's death at the altar are causally linked, Priam's death being the culmination of the series of events set in motion by the sacrifice at Aulis. Lucretius' motive for invoking Priam's death at 1.84–5 is thus clear: his rejection of irrational variants of the Iphigenia story is further underlined via allusive reference to the kind of uncontroversial death that is easily integrated into a rationalized Epicurean view of mythic history; what is more, the uncontroversial death in question is both causally related to and visually reminiscent of the death of Iphigenia at Aulis.Footnote ⁵⁰

Like Accius' Clutemestra, there is good reason to believe that Ennius' Andromacha would have been a familiar part of the dramatic repertoire in the fifties b.c.: it is among the fragments quoted by Cicero in the pro Sestio (121), reporting the performance of Aesopus at the ludi Apollinares of 57. It is possible that the play was performed in its entirety; it is likewise possible that quotations from it were inserted by Aesopus into a different play (probably Accius' Eurysaces).Footnote ⁵¹ Either way, it is reasonable to assume that a culturally competent Roman reader of the type anticipated in DRN would be familiar with this play (or at least with the famous scene in question).Footnote ⁵²

Lucretius' allusive strategy at 1.84–100 is significantly more complex than the straightforward demythologization of Accius' Clutemestra at 6.387–95. In the case of Iphigenia, the tragic genre offers a polyphonic tradition including both rational and irrational variants. Affinities with the story told in Aeschylus' Agamemnon (from the rational branch of the tradition) help Lucretius to demythologize the alternative account found in Euripides' IA. Different tragedies are thus played off against one another, with one contributing to the demythologization of the other. Lucretius' rationalizing assertions are further reinforced via the invocation of an uncontroversial death from another tragic text (Ennius' Andromacha), which would have been particularly salient for theatre-going readers in the mid first century b.c. As such, it is not possible in this instance to identify a single overarching Lucretian attitude to the explanatory value of tragedy as a genre. The mythic content of tragedies can be true or false, rational or irrational, with the result that the explanatory power of tragedies may vary from play to play. In such cases, therefore, the proper response to the mythic content of what we experience in the theatre is not outright rejection, nor outright acceptance, but rather selective ambivalence—careful and considered selection of those elements of the tradition that can be placed into alignment with the basic principles of Epicurean rationalism, with simultaneous rejection of other elements. Lucretius' implicit advice to the reader here is neither to reject nor to accept the lessons of tragedy outright, but rather to attend the theatre with open eyes.

DRN 5.318–23: PACUVIUS' CHRYSES

Among the extant fragments of Chryses we find evidence of a complex Pacuvian adaptation of Euripides, fr. 839 Nauck (from Chrysippus)—a text which would later be rendered into Latin (with some important changes)Footnote ⁵³ by Lucretius himself, at 2.991–1001.Footnote ⁵⁴ Given the very large amount of critical notice that Lucretius' rendering of the Euripides fragment has received,Footnote ⁵⁵ I have chosen here to focus my attention on the closely related but less well-studied allusion to Chryses at 5.318–23,Footnote ⁵⁶ which forms the conclusion to a series of arguments for the mortality of the earth and its constituent parts:

If the sky creates all things and receives them again upon death (that is, if it is both diminished and replenished), then it is mortal (here we should understand the unsupplied additional premiss that everything subject to change is mortal).Footnote ⁵⁷ In such an argument we might normally expect a minor premiss affirming the content of the protasis (‘the sky does create and receive all things') leading, via modus ponens, to the conclusion that the sky is indeed mortal. But no such premiss is stated, and the distancing phrase quod quidam memorant at 320 leads us to suspect that Lucretius is not in fact committed to the truth of the protasis at all. Indeed, as the commentators point out, these lines are out of keeping with the ‘Euripidean’ passage at 2.991–1001, where the earth, ‘impregnated’ by the sky, is herself primarily responsible for the creation of life (994–7), receiving earth-born creatures after death (999–1000; cf. 5.259), with the produce of the sky returning to the sky (1000–1). So, we are left with an unusual situation in which Lucretius bases an argument on a premiss which, given what he states elsewhere in the poem, he ought not to accept, but which presumably is still understood to have persuasive force. In order to consider this problem further we must first take a deeper look at the underlying dramatic text. The passage 5.318–23 closely recalls Pacuvius, Trag. 86–7 R³ (= fr. 79.1–2 Schierl) and Trag. 90–2 R³ (= fr. 80 Schierl):

These form part of a set of philosophically themed fragments attributed to Pacuvius' Chryses, which, although quoted in quite different contexts, are none the less sufficiently similar in content to be considered elements of a cohesive whole.Footnote ⁵⁸ While the influence of Euripides' Chrysippus on Pacuvius' phrasing is clear, the content has changed: here the sky enjoys a greatly enlarged role, with the role of the earth reduced to the birthing (parit) of bodies (Trag. 93 R³ = fr. 81 Schierl), whose creation, animation, shaping and supplementation are all attributed to the sky.Footnote ⁵⁹ Trag. 90–2 R³ are from one of two fragments preserved by Cicero at de Divinatione 1.131, which, Cicero tells us, are spoken by the same characterFootnote ⁶⁰ —a physicus Footnote ⁶¹ —probably in the same scene (cum ipse paucis interpositis uersibus dicas …). The first of these two fragments (Trag. 83–5 R³ = fr. 77 Schierl), which may also have been familiar to Lucretius (compare Trag. 84 R³ ex alieno iecore sapiunt to DRN 5.1131–2 sapiunt alieno ex ore), contains a rejection of ornithomancy and hepatoscopy. When these rationalizing philosophizing fragments spoken by the physicus are contrasted with those fragments of the same tragedy that feature ‘highly Romanized religious language’,Footnote ⁶² a picture soon emerges of a play in which a confrontation—perhaps even a formal agōn—is enacted between traditional Roman religion and ‘naturalistic’ philosophical attitudes (a confrontation which is enabled and activated by sustained allusion to the most well-known example of Euripidean philosophizing).Footnote ⁶³ I suggest that the existence of this confrontation in Chryses provides the key to the interpretation of DRN 5.318–23.

We may start by considering the referent of quidam at 5.320. Previous interpretersFootnote ⁶⁴ have understood this as an oblique reference to the Stoics, with elements of whose cosmology the content of 5.318–23 has been said to bear some affinities. There are two problems with this approach. First, the ‘affinities' in question are remarkably unspecific; the content of the protasis at 319–20 may be thought of as Stoic only in the sense that it is broadly consistent with a Stoic view. We have no reason here to pin down the referents of quidam at 320 to Stoics and Stoics alone.Footnote ⁶⁵ Second, and more importantly, there is no need to invoke the Stoics as the implicit referents of quidam, because we already happen to know precisely who says the kind of things mentioned here, namely the physicus of Pacuvius' Chryses. Rather than thinking of the Pacuvian fragments as an oblique way for Lucretius to get at the Stoics, we should think of the contents of those fragments as philosophical in their own right—as an example of the kind of generic philosophizing which Roman readers would regularly have encountered in a variety of cultural domains (poetry, historiography or, as in this case, at the theatre). As such, 5.318–23 should not be understood as containing an indirect allusion to a particular philosophical doctrine or school (such an interpretation being born of too narrow a conception of how philosophical discourse was experienced in Republican Rome), but should rather be taken at face value, as a direct allusion to the theatrical text from which it is sourced—Chryses—with quod quidam memorant functioning as a programmatic marker alerting the reader to the presence of the Pacuvian intertext.

Gale (n. 22), 183–4 identifies this passage as an example of ‘latent myth’, stating that ‘rather neatly, Lucretius uses the myth to prove that, far from being divine, the aether must in fact be perishable’ (184). Thus Gale sets up an opposition between the content of the ‘myth’ underlying 5.318–23 (by which is meant the marriage of earth and sky in a hieros gamos) and Lucretius' rationalizing treatment of it: while the ‘myth’ asserts the divinity of aether, Lucretius' treatment of the ‘myth’ includes the rejection of said divinity. We should note, however, that it is by no means certain that the original context in which the natural-philosophical fragments of Chryses were found involved any commitment whatever to the actual divinity of earth and sky. As we have seen, in Chryses Pacuvius enacted a dialogic confrontation between traditional religion and naturalistic philosophy. It is unlikely, therefore, that the physicus who speaks these lines would have put forward a naïve or uncritical account of the mythical divinity of earth and sky. Which is to say, the imagery of the hieros gamos was most likely already a case of latent myth in its original philosophizing context in Chryses. Rather than seeing Lucretius at 5.318–23 as being engaged in demythologization of the kind envisaged by Gale, I suggest that we should understand these lines as appropriating an existing piece of well-known philosophical allegory, removing it from the dialogic context of Pacuvius' Chryses and relocating it to the fully rationalized univocal environment of DRN. While the content of the protasis at 5.319–20 is, strictly speaking, not in alignment with claims made elsewhere in DRN (hence the phrase quod quidam memorant and the lack of any minor premiss to the argument), Lucretius is, none the less, keen to appropriate this small piece of familiar theatrical philosophizing in order to provide additional force in favour of his main argument (that the earth and all its constituent parts are mortal). Recall that this is only the last in a sequence of arguments to the same effect: the main point is already proved along Epicurean lines; the popular philosophizing of Roman tragedy is brought in only at the end of the sequence. Nevertheless, we can see once again that the kind of dramatic content to which readers were regularly exposed at the theatre is not uniformly antithetical to Lucretian rationalism. In Pacuvius' Chryses, the dialogic opportunities offered by the genre of tragedy were exploited in order to enact a confrontation between two basic modes of thought, the one religious, the other philosophical. Lucretius, in his reception of that text, appropriates the content of just one half of the dichotomy and so implicitly rejects the other. Once again, the proper reaction to such a play among readers of DRN is neither outright rejection nor outright acceptance, but rather the careful selection of those aspects of the theatrical intertext that are most well suited to the Epicurean rationalist perspective.