1. LUCIAN

Who exactly are these νόθοι καὶ παρέγγραπτοι, forced out by Momus? At the beginning of Lucian’s Concilium deorum, Hermes announces that the divine assembly will concern ‘metics and foreigners’ (περὶ τῶν μετοίκων καὶ ξένων, Conc. D. 1). Momus’ first complaint involves Dionysus, who is ‘half human and not even a Greek on his mother’s side, but the grandson of some Syro-Phoenician traveller named Cadmus’ (ἡμιάνθρωπος ὤν, οὐδὲ Ἕλλην μητρόθεν ἀλλὰ Συροφοίνικός τινος ἐμπόρου τοῦ Κάδμου θυγατριδοῦς, 4). This description shows that Dionysus is a νόθος in the divine polis in two senses. Since Lucian’s divine polis is a Greek polis modelled on (post-)Periclean Athens, he is a bastard in a sense that would have been most common in the human polis: he has a foreign and hence non-citizen mother.Footnote ³ At the same time, Lucian uses the notion of bastardy in an extended, ontological sense that can only apply to the divine polis: Dionysus is a ἡμιάνθρωπος, a son of a mortal mother. That this second sense is crucial is confirmed when Zeus, having an inkling of where the argument is headed, warns Momus not to go after Asclepius and Heracles next (6). Here, the foreignness of the mother is no longer an issue: what Dionysus, Asclepius and Heracles share and what makes them bastards is that they have a divine father and a mortal mother.Footnote ⁴ This ontological sense of bastardy remains salient when Momus accuses Zeus of being responsible for the bastardization (τοῦ νοθευθῆναι) of the divine assembly when he, a god, started sleeping around with mortal women (θνηταῖς, 7). At this point, it has become clear that the same ontological meaning also applies to the label μέτοικοι καὶ ξένοι, which Hermes used at the very beginning: the meeting on Olympus concerns non-divine subjects now living among the gods. These newcomers falsely claim citizenship privileges in the divine polis and are thus παρέγγραπτοι: their names have been illegally added to the citizen lists.Footnote ⁵ The notion of νόθοι καὶ παρέγγραπτοι, then, brings together questionable descent and administrative illegality.Footnote ⁶

This connection between bastardy, illegal registration and divine status is not Lucian’s invention. It occurs earlier in Aristophanes’ Birds, where an embassy consisting of Poseidon, Heracles and a foreign god are sent to negotiate a settlement with the birds, who have blocked the passage of sacrifices from humans to gods.Footnote ⁷ Surprisingly, perhaps, the accusation of bastardy is levelled not against the foreign god but against Heracles. As in Lucian, then, the familiar notion of bastardy is used to describe ontological status: it does not bring across a point about foreign descent but about non-divine descent. Poseidon points out that Heracles will not have his share of his father Zeus’s inheritance because he is a νόθος (Au. 1650, ed. OCT): his mother is foreign (ὤν γ’ ἐκ ξένης γυναικός, 1652).Footnote ⁸ Solon’s law on inheritance is applied to the divine realm to disenfranchise him (1660–6). Poseidon tacks on a second argument: knowing that the answer is ‘no’, he asks if Zeus has inducted Heracles into his phratry (εἰς τοὺς φράτερας, 1669). This mirrors Lucian’s addition of παρέγγραπτος to νόθος as it connects bastardy to citizenship lists.Footnote ⁹ As a scholium to this verse points out (using the verb ἐγγράφω), Aristophanes ‘again comically transposes the practices of the Athenians to the gods. The Athenians considered the enrolment in the phratries a sign of legitimate birth’ (πάλιν ὡς ἐν κωμῳδίᾳ μετήγαγε τὰ Ἀθηναίων ἔθη ἐπὶ τοὺς θεούς. διὰ γὰρ τοῦ ἐγγραφῆναι εἰς τὰς φατρίας σύμβολον εἶχον τῆς εὐγενείας οἱ Ἀθηναῖοι).Footnote ¹⁰ Like Lucian, Aristophanes applies human political practices to the divine realm through a similarly extended ontological use of νόθος and ξένος.Footnote ¹¹

What is at stake in Lucian’s dialogue, however, is not the status of a single god or even a well-defined group of gods. Momus’ main concern is that, if bastards are illegally admitted to the divine polis, it will be impossible to put a stop to the admission of new gods of all sorts. At first, he foregoes Dionysus’ own dubious status because this son of mortal Semele has been granted immortality by an earlier vote of the divine assembly (Conc. D. 4). Instead, Momus accuses Dionysus indirectly by going after the many new deities the god of wine has brought to Olympus as his posse (Pan, satyrs, Ariadne, Erigone and her dog, 5). This reveals a slippery slope: if the half-human Dionysus can be a god, it is a small step to the deification of his friends and lovers and even their pets. Or again, if Heracles is a god, there is no good reason why his master Eurystheus is not. And if Dionysus is a god, the exclusion of his cousins Pentheus, Actaeon and Learchus is hard to explain (7). Eventually, Zeus admits the risk of this snowball effect: ‘it is necessary to put a stop to most of these issues, lest they increase’ (δεῖ τὰ πολλὰ αὐτῶν ἐπισχεῖν, ὡς μὴ ἐπὶ πλεῖον ἂν γίγνηται, 14).Footnote ¹²

2. CICERO

Like the application of bastardy and illegal registration to the divine realm, the use of slippery-slope reasoning to paint the picture of an overcrowded Olympus is not of Lucian’s own making. Cicero (Nat. D. 3.43–52) and Sextus Empiricus (Math. 9.182–90) report a similar series of arguments, which they attribute to Carneades, one of the heads of the Platonic Academy in its sceptical phase. Sextus Empiricus notes that these arguments are called σωρίτης (from σωρός, ‘heap’) arguments.Footnote ¹³ They point out the vagueness of certain concepts such as heaps. Is one grain of sand a heap? Obviously not. Are two or three or even ten grains? Still no, but we start to see how we will end up with an undisputed heap of sand at some point. Which exact grain has caused the answer to shift from no to yes? Or, if we begin with an undisputed heap and start picking away grains one by one, from yes to no? It seems impossible to draw the line. Carneades applied this to the concept ‘god’ by proposing chain arguments that seem to start out with a firm yes (is Zeus a god?) and to end with a firm no (is a mountain stream a god?), thus exposing a problem of demarcation.

The argument that goes from Zeus to the mountain stream, as reported by Sextus Empiricus (Math. 9.182–3), uses the notion of brotherhood to jump from Zeus’s divinity to Poseidon’s and thus projects family relations onto the divine realm. Similarly, when we turn to Cicero’s report of Carneades’ arguments, we find νόθοι once again implicated in a slippery-slope argument. The character Cotta, Cicero’s literary evocation of the orator and consul whom he has representing the sceptical Academy, is talking (Nat. D. 3.45):

This passage ends with the explicit and humorously provocative application of the ius ciuile to the ius naturae of divine status, which resembles Lucian’s (and Aristophanes’) practice, albeit fitted to the Roman context. Cicero begins the chain argument with undisputed gods such as Apollo, Hephaestus and Hermes. He ends with figures such as Aristaeus and Theseus, whom no one would have considered gods. As always, the question of the sorites is: where is the cut-off? This is a particularly urgent question given the in-between category that includes Heracles, Asclepius, Dionysus (identified with Liber), and Castor and Pollux. The trio encountered earlier in Lucian is now joined by the Dioscuri, who are also counted among the traditionally recognized deities with a mortal mother.Footnote ¹⁵

Although Lucian and Cicero share several elements (the sorites-style argument applied to divine status, the invocation of civil law and family connections, the examples given), the motif of bastardy and illegal registration is not explicitly present in the latter. I submit, however, that we should connect Nat. D. 3.45 (quoted above), which is part of the series of sorites arguments explicitly attributed to Carneades (3.43–52), with Nat. D. 3.39. The latter paragraph should be read as the character Cotta’s rhetorical introduction to these arguments. He compares Stoic theology to ‘the ignorance of the naive masses’ (uolgi atque imperitorum inscitiam). The point of comparison is that both groups recognize a staggering multitude of gods (quanta autem est eorum multitudo: mihi quidem sane multi uidentur, 3.40).Footnote ¹⁶ While this again vaguely recalls Momus’ complaint about overcrowding in Lucian, Cicero’s description of how the naive masses conceive of the gods renders the parallel more tangible (Nat. D. 3.39, ed. Teubner, modified):

A few paragraphs later, as we have seen, Cotta will group Heracles, Asclepius, Dionysus, and Castor and Pollux together as examples of those who are considered gods, although they have mortal mothers (dei sunt habendi mortalibus nati matribus) and use them in a sorites argument. In the present passage, which comes before the formal sorites arguments, he introduces not only that group but also two other groups that will feature in those arguments. (1) Syrian and Egyptian animal worship foreshadows the following sorites argument: if the traditional gods (illi quos colimus et accepimus) are really gods, then so must be Serapis and Isis; if the latter, then also gods of the barbarians (barbarorum deos), which means that a whole host of animals (dogs and fishes are included in the enumeration) will have to be counted as gods (Nat. D. 3.47). (2) Alabandus and Leucothea return as well: if Leucothea is a goddess, on account of her being the daughter of (Zeus’s descendant) Cadmus, then so must be (mortals such as) the seers Amphiaraus and Trophonius and hence also the subjects of hero cults such as Alabandus (Nat. D. 3.49–50).Footnote ¹⁸ These parallels strongly suggest that the repeated mention of Heracles, Asclepius, and Castor and Pollux in Nat. D. 3.39 (on the religion of the ignorant mob) and 3.45 (the sorites arguments) is not a coincidence and that we can discern the sorites logic already in 3.39.

As we have seen, Heracles cum suis are labelled mortalibus nati matribus at 3.45, with reference to the ius ciuile on such matters. The currently accepted text of 3.39 describes them as nouos et adscripticios ciues. Footnote ¹⁹ I suspect that a problem with the text obscures Cotta’s specific point as well as the connection between De natura deorum and the other works under discussion. The scribes surely had trouble with the word adscripticios: several of the primary manuscripts have a variant. As Pease remarks, this adjective ‘is very rare save in the legal codices’.Footnote ²⁰ While the manuscripts agree on nouos, the words nouos et adscripticios do not capture the specificity of this group as mortalibus nati matribus, which sets them apart from the group of Leucothea and Alabandus: a claim to divinity should be denied on the grounds of bastardy if the applicant’s mother is not a god.Footnote ²¹ That this is Cotta’s point is confirmed rather than denied by the fact that Romulus joins the group of nothi at Nat. D. 3.39, while Liber/Dionysus is added at 3.45: both are heavenly bastards with a divine father and a mortal mother.Footnote ²² To capture the legal concept of bastardy in one adjective, the Latin language had to transcribe the Greek word νόθος, as Quintilian points out.Footnote ²³ Hence my proposal to read nothos et adscripticios. The rare form nothos probably would have caused the scribes just as much trouble as adscripticios, but is more readily changed to a familiar word, as is suggested by another passage from Cicero: in QFr. 2.12.4 Shackleton Bailey similarly proposed to emend the manuscripts’ notum to not<h>um.Footnote ²⁴

In addition to the general argumentative parallels with Lucian indicated earlier, then, Cicero uses the specific motif of bastardy and illegal registration in his take on the theological sorites arguments after all: he introduces a calque of νόθοι καὶ παρέγγραπτοι (as Lucian phrased it) into Latin by describing Heracles, Asclepius, Castor and Pollux, and Romulus as nothi et adscripticii ciues. But to what end? The parallels established earlier between the attack on ‘the vulgar and the ignorant’ (3.39) and the formal sorites arguments aimed at the Stoics (3.45) underscore the point that the Stoics are just as bad as the ignorant mob (3.40): neither group can come up with a consistent theological account. Sextus Empiricus, after presenting his version of the theological sorites arguments attributed to Carneades, similarly compares dogmatic philosophers to the ignorant mob (Math. 9.191–2). This may add to the suspicion that Cic. Nat. D. 3.39, although it comes before the sorites arguments stricto sensu, was already inspired by Carneadean scepticism.Footnote ²⁵

3. PLUTARCH

Cicero’s character Cotta thus chides the rational theology of the Stoics along with the popular religion of the mob. He considers both to be a threat to the traditional religion that he himself as a pontifex endorses both at the beginning of his speech (Nat. D. 3.5–6) and when embarking upon the sorites arguments (3.43). How Cotta could combine this traditional religious piety with his theological scepticism has been the subject of debate.Footnote ²⁶ The De natura deorum itself does not give us much to resolve this issue, since Cotta’s contribution is concerned almost exclusively with refuting the positions of the other characters (cf. Nat. D. 1.57–61). Hence, I propose to approach this debate from a new angle by comparing the character Cotta with a figure who similarly combines a penchant for scepticism in theology with a priesthood and a firm devotion to traditional religion: the Platonist Plutarch of Chaeronea who, like Cicero, enjoyed casting himself as a character in his dialogues.Footnote ²⁷ In the dialogue Amatorius, the character Plutarch comes to the defence of Eros, whom he considers the oldest and greatest god (Amat. 756C, ed. Teubner, modified):

I have quoted Hubert’s Teubner text with one modification: I propose to change the manuscripts’ παρεισγραφῆς to παρεγγραφῆς. The word παρεισγραφή, which occurs in both manuscripts, is recorded as a hapax legomenon in major dictionaries and is printed in all editions.Footnote ²⁸ There is no trace of a related adjective or verb. From the context it is sufficiently clear that this hapax legomenon would designate the crime connected to the act of παρεγγράφειν, which I have been discussing: the act of illegally adding someone to a list of citizens. My suggestion, παρεγγραφή, is not attested in Greek literature and is absent from dictionaries. The LSJ Supplement, however, points to an attestation of the word: a stone found in Lycia contains an edict about corruption in urban archives that tells about a public slave being punished for accepting ‘lists that contain fraudulent additions and omissions’ (παρενγραφὰς κ̣[αὶ ἀ]παλοιφὰς ἔχοντα πιττάκια, SEG 33.1177.10–11).Footnote ²⁹ The edict was issued in 43 c.e.—that is, around the time of Plutarch’s birth—which renders it plausible that the word was current in Plutarch’s time. Regardless, παρεγγραφή is the obvious substantive to go with the well-attested words παρεγγράφω, παρέγγραπτος and παρέγγραφος, and makes more sense than παρεισγραφή. Moreover, the word ἐγγραφή is well attested to denote the legal addition of someone to a citizen list.Footnote ³⁰ Prefixing παρα- to signal wrongfulness is a small step. The corruption from παρεγγραφή to παρεισγραφή may have occurred either in the uncial stage of the transmission (uncial gamma being easily confused with uncial sigma)Footnote ³¹ or later, when παρεισγράφω entered the lexicon and perhaps became more familiar than παρεγγράφω.Footnote ³²

It is beyond doubt that Plutarch knew the correct legal terminology involving illegal administrative additions (that is, παρεγγραφ- and not παρεισγραφ-), as he uses it elsewhere (cf. Vit. Ant. 15.3). Moreover, he also combined this terminology with the notion of bastardy elsewhere. In Quaest. conu. 7.6 the character Plutarch and his friends are discussing the sympotic phenomenon of ‘shadows’, uninvited dinner guests brought along by invited guests. One of the interlocutors remarks that one will never be at ease at a symposium if one is, ‘as it were, a bastard illegally registered at the symposium’ (τρόπον τινὰ νόθος καὶ παρεγγεγραμμένος εἰς τὸ συμπόσιον, Quaest. conu. 7.6.707E–F, ed. Teubner). The metaphorical application of the legal notions of bastardy and illegal registration recalls the passage from the Amatorius.

These observations suggest that we should read παρεγγραφῆς instead of παρεισγραφῆς and that the supposed hapax legomenon παρεισγραφή should be omitted from our dictionaries. If this is granted, the lexical connection between Plutarch’s text and that of Lucian discussed earlier (and, hence, also with the texts by Aristophanes and Cicero under discussion) becomes clearer than current editions let on.

The gods whom Plutarch accuses of νοθεία and παρεγγραφή are called Ἄτται τινὲς καὶ Ἀδωναῖοι. What kind of gods does Plutarch have in mind? Ἄτται must be entities such as Attis, the self-castrated lover of the Phrygian mother-goddess Cybele. Cybele (known as Magna Mater in the Roman empire) and her deified companion Attis rose to prominence in Plutarch’s time.Footnote ³³ Pausanias (7.17.10–13), who wrote a few decades after Plutarch’s death, reports that there are various versions of the Attis myth, but Plutarch’s take on Attis’ divine status is clear enough from a digression on myths involving marriages between a mortal and a god in the Life of Numa. The digression is sparked by Numa’s alleged involvement with the nymph Egeria. Plutarch comments that this story resembles the Phrygian myth about Attis and, while insisting on the general philanthropy of the gods, concludes: ὡς δὲ καὶ σώματος ἀνθρωπίνου καὶ ὥρας ἐστί τις θεῷ καὶ δαίμονι κοινωνία καὶ χάρις, ἔργον ἤδη καὶ τοῦτο πεισθῆναι (‘That a god or demon would derive any experience of communion or carnal pleasure from a human body and its beauty, this, surely, is hard to believe’, 4.5, ed. Teubner; transl. B. Perin in the LCL series). Attis, then, is placed squarely in the human realm and so must be the Ἄτται that Plutarch has in mind.

The Ἀδωναῖοι are more elusive. While Hubert’s Teubner edition follows the manuscripts in printing it, other editors adopt Bergk’s conjecture Ἀδώνιοι.Footnote ³⁴ Most major dictionaries similarly cite the passage at hand as reading Ἀδώνιοι and do not include Ἀδωναῖος as a lemma.Footnote ³⁵ In Plutarch’s mind, however, it probably made sense to use Ἀδωναῖος in order to suggest a connection between Adonis and the Jewish god addressed as Adonai, whose possible identification with Dionysus is mentioned in Plutarch’s Quaestiones conuiuales.Footnote ³⁶ While, then, the category of Ἀδωναῖοι certainly includes Adonis, who—like Attis—was considered a deified mortal, we have to consider the possibility that Plutarch, like Lucian and Cicero, also implicates Dionysus in the charge of νοθεία and παρεγγραφή.Footnote ³⁷

This becomes clear if we consider how Plutarch views the possibility of marking off the realm of gods in a digression in his Life of Pelopidas. He is talking about the surroundings of the temple of Apollo at Tegyra (Vit. Pel. 16.6–8, ed. Teubner):

While Apollo is a θεός because he is unborn, Heracles and Dionysus are δαίμονες.Footnote ³⁸ Here, then, Plutarch draws a clear line between gods and non-gods, with Heracles and Dionysus excluded from the former category, by refusing any transition from non-god to god or any contamination of divine nature by non-divine nature.Footnote ³⁹ Plutarch’s solution to the challenge posed by the sorites is to rely on ὁ πάτριος λόγος as evidence (τεκμαίρεσθαι). The same solution is put forward in the context of the passage from the Amatorius under discussion: right before the passage quoted above, the character Plutarch rebukes the demand for proof (λόγον … καὶ ἀπόδειξιν) about each god individually (περὶ ἑκάστου). Instead, he advises to rely on ἡ πάτριος καὶ παλαιὰ πίστις as the sole τεκμήριον and accept the traditional pantheon as a whole (Amat. 756B).

This solution to the problem posed by the sorites, then, is not a refutation. On the contrary, like Cicero’s Cotta, Plutarch affirms the logic of the sorites and its force in matters of rational theology. Theological reasoning, which asks for a περὶ ἑκάστου τεκμήριον in the form of λόγος καὶ ἀπόδειξις, cannot put a stop to the sorites: if one god is put under scrutiny like that, the whole pantheon falls like a row of dominoes as the sorites renders making a cut-off between gods and non-gods impossible.Footnote ⁴⁰ Instead, ἡ πάτριος καὶ παλαιὰ πίστις is the only τεκμήριον in matters concerning the gods (Amat. 756B). Similarly, Cotta criticizes the Stoics for trying to prove the existence of the gods with many arguments (multis argumentis), while he himself relies on ancestral tradition as the only argument (mihi enim unum sat erat, ita nobis maioris nostros tradidisse, Nat. D. 3.9; cf. 3.6). Plutarch and Cotta agree: traditional religion, not rational theology, provides a sure way of engaging with the divine. They do not win the game against the sorites. They simply refuse to play.

While Lucian and Cotta involve the νόθοι καὶ παρέγγραπτοι in a sceptical argument that brings out the impossibility of categorically distinguishing gods from non-gods and thus the impossibility of excluding the νόθοι καὶ παρέγγραπτοι, Plutarch invokes the νόθοι καὶ παρέγγραπτοι in a seemingly opposite way: in the Amatorius, he confidently draws the line between Eros and the νόθοι καὶ παρέγγραπτοι to the exclusion of the latter. Both reasonings are sides of the same sceptical coin: in the rational theological sphere the line cannot be drawn, but in the traditional religious sphere it can. Plutarch’s brand of scepticism fully recognizes that matters of theology are beyond proof.Footnote ⁴¹ Like Cotta, he relies on tradition instead of reasoning when engaging with the divine. Combined, the characters Cotta and Plutarch show how the νόθοι καὶ παρέγγραπτοι can be part of both the theological sorites argument (which makes distinguishing between gods and non-gods impossible) and the religious immunization strategy against it (which makes distinguishing between gods and non-gods possible again). Before finally returning to Lucian, I note two complications in Plutarch’s use of the νόθοι καὶ παρέγγραπτοι that will eventually give us better insight into Lucian’s work and into the workings of the νόθοι καὶ παρέγγραπτοι argument in general.

First, it should be noted that Plutarch, by choosing Attis and Adonis as representatives of the νόθοι καὶ παρέγγραπτοι, does not insist on the pattern of divine father and mortal mother found in Lucian and Cicero. Although they can certainly be considered παρέγγραπτοι in the divine polis (given their human ontological status), Attis and Adonis are not clearly divine νόθοι (bastards with a divine father and a mortal mother).Footnote ⁴² The combination of νοθεία and παρεγγραφή may have resonated as a formula for Plutarch. The fact that he uses the formula while broadening its meaning suggests that he was familiar with a broader set of Carneadean sorites paradoxes of the sort that appears in Cicero and Sextus Empiricus and, in a less formal fashion, in Lucian.Footnote ⁴³ The common denominator that Plutarch has in mind could simply be the presence of a contaminating trace of humanity in the divine pretenders who should be excluded. Against this class of sorites arguments his religious scepticism can provide immunity by demarcating gods from non-gods, not through argument but through tradition.

Second, the context in which Plutarch’s mention of the νόθοι καὶ παρέγγραπτοι appears suggests that we should not consider only part-humans to be intruding into the divine realm but also abstract concepts. The character Plutarch’s appeal to the πάτριος καὶ παλαιὰ πίστις that manages to exclude νόθοι καὶ παρέγγραπτοι from the divine realm is a response to another character’s attack on the divinity of the god Eros. That character holds that Eros is a deified πάθος (Amat. 755E–F). That Plutarch’s response to this attack consists in affirming the exclusion of those liable to νοθεία and παρεγγραφή is perhaps more readily understood if we consider that Eros as a deified πάθος figures in one of Carneades’ theological sorites arguments (as reported by Sext. Emp. Math. 9.187): if Aphrodite is a goddess, then so is her son Eros, who is a πάθος, and hence all other πάθη such as fear.

The Stoics must have appeared as vulnerable to this sorites as to the one involving νόθοι καὶ παρέγγραπτοι. A doxographic passage usually taken to represent Stoic doctrine includes deified πάθη, among which Eros is listed, as a category of gods (Aët. 1.6 = SVF 2.1009). In a passage drawing on the same material as the doxography, Balbus, Cicero’s Stoic character, similarly includes Cupido among the forces deified on account of their overwhelming force. He calls Cupido a uitium but does so without questioning his divinity (Cic. Nat. D. 2.60–2).Footnote ⁴⁴ Deified πάθη seem to be an outgrowth of the Stoic practice of deifying abstractions—an outgrowth as unwanted and as hard to evade as the inclusion of any human among the gods, as the sorites arguments point out.Footnote ⁴⁵

The character Plutarch’s insistence on tradition rather than argument, on the other hand, serves to immunize against the sorites logic in all its manifestations: asserting that Eros is a true god amounts to asserting that he is neither a νόθος καὶ παρέγγραπτος nor a πάθος. Like the previous complication, this again leads to the suspicion that Plutarch’s brisk mention of νοθεία and παρεγγραφή in the Amatorius was part of a larger sceptical body of argument that recognized the devastating results of the sorites on a rational theological level and then sought immunity from these results on a traditional religious level.Footnote ⁴⁶

4. BACK TO LUCIAN

In Lucian’s Concilium deorum, Momus wanted to rid Olympus of νόθοι καὶ παρέγγραπτοι such as Dionysus, Asclepius and Heracles. Cicero’s De natura deorum showed that Lucian’s slippery-slope reasoning about the inclusion of νόθοι καὶ παρέγγραπτοι goes back to Carneades’ theological sorites arguments. Plutarch’s Amatorius gave further hints as to how this Carneadean scepticism in theological matters could be combined with traditional religion, perhaps in line with how Cicero’s Cotta conceived of it. While using the formulaic conjunction of both νοθεία and παρεγγραφή, Plutarch seems to have broadened its scope to include not only the sons of divine fathers and mortal mothers but more generally those who are tainted by humanity as well as deified abstractions such as πάθη—two categories that were also included in the broader set of sorites arguments attributed to Carneades.

Lucian’s Concilium deorum similarly extends the charge of νοθεία and παρεγγραφή beyond the strict definition, established early in the work, of which Dionysus, Asclepius and Heracles are the prototypical examples (Conc. D. 2–8). Like Cicero, Lucian notes that there is not only the standard pattern of divine father and human mother but also the opposite combination (8). Next, he turns to foreign candidates for divinity (9–11). As in Plutarch, Attis is accused.Footnote ⁴⁷ Like Cicero’s Cotta, who associated foreign religion with animal worship, Lucian’s Momus is quick to point out that the Egyptian god at the table is a doghead. Then comes a pet peeve of Lucian’s: the ubiquity of fraudulent oracles.Footnote ⁴⁸ Trophonius and Amphilochus are singled out as oracle mongers. Again, Cicero provides a parallel by involving Trophonius and Amphiaraus in a sorites argument, as we have seen. The substitution of Amphilochus for his family member Amphiaraus may well be due to Lucian’s personal interest in the former: Lucian mentions Amphilochus as an inspiration for Alexander, the false prophet to whom he devoted a work (Alex. 19).Footnote ⁴⁹

At this point, Momus announces the conclusion of his treatment of οἱ νόθοι καὶ παρέγγραπτοι (Conc. D. 13), the definition of which has been broadened as in Plutarch. However, instead of immediately reading his decree, which will require divine candidates to provide proof of parentage, he tacks on a rant about deified abstractions such as ἀρετή, φύσις, εἱμαρμένη and τύχη. These are not present in the assembly since, as Momus points out, no one has ever seen them. As such, and again as in Plutarch, the involvement of abstractions may seem ill-fitted to the context of νόθοι καὶ παρέγγραπτοι. However, if we consider both the νόθοι καὶ παρέγγραπτοι and the deified abstractions under the umbrella of the sorites arguments, aimed at champions of rational theology such as the Stoics, this step makes more sense.

It has become clear that from Cotta and Plutarch we can piece together an Academic sceptic who invokes sorites arguments against rational theology and turns instead to traditional religion, which appears immune to such arguments. But to what use is Lucian putting the theological sorites? Not unlike Cotta and Plutarch, Lucian’s Momus appears to use traditional religion as an immunization strategy against the sorites: his decree regards the inclusion of illegitimate pretenders along with the ancient and true gods (τοὺς παλαιούς τε καὶ ἀληθεῖς θεούς) as going against ancestral tradition (παρὰ πάντα τὰ πάτρια, Conc. D. 14). The attempt to rehabilitate this tradition recalls Plutarch’s insistence on ἡ πάτριος καὶ παλαιὰ πίστις (Amat. 756B) and Cotta’s appeal to the mos maiorum (Nat. D. 3.43).

As it turns out, however, Momus does not succeed in drawing the line as clearly as Cotta and Plutarch: his efforts turn out to be self-defeating and his immunization strategy ultimately fails, which contributes to the humorous nature of the work. After explaining his strategy of having the candidates for divinity proving their lineage, he adds that, among the remaining true gods, each one should have their own single speciality: Athena should stop curing people, Asclepius should refrain from giving oracles, and Apollo should cease being such a busybody altogether (Conc. D. 16). After being grouped among the νόθοι (6), Asclepius is now envisaged as a permanent resident of Olympus. Momus, then, does not stick to his own criteria.Footnote ⁵⁰

This could well be for reasons of self-preservation. After all, how justified is Momus’ own presence in the divine assembly? He seems to be implicated by one of the Carneadean sorites arguments as reported by Cicero (Nat. D. 3.44): through a genealogical chain, the argument runs from Jupiter (definitely a god) to the children of Night, monstra among which Momus should be counted (definitely not gods). The list of children of Night, which closely follows that in Hesiod’s Theogony (211–25), includes Amor, Inuidentia, Miseria, Querella, Gratia, Pertinacia and thus resembles in a way the use of deified πάθη in Sextus Empiricus’ report of the sorites arguments.Footnote ⁵¹

It looks like Momus’ decree will not solve the overcrowding of Olympus after all: the very decree that was supposed to banish the νόθοι καὶ παρέγγραπτοι assumes the continued presence of a νόθος (Asclepius) among the gods and was written by a πάθος. By proposing a principle and then immediately forsaking it, Momus appears all too human, as Lucian’s gods tend to do.Footnote ⁵² Whatever may be Lucian’s own agenda, this story of self-defeat, which presupposes an understanding of the sorites logic at play, is essential to the humour of the piece.

5. CONCLUDING REMARKS

As Cicero’s Cotta shows, the νόθοι καὶ παρέγγραπτοι claiming divinity present a problem for a rational theology. Logically, they can be neither included among nor excluded from the gods. As sorites paradoxes show, it is impossible to draw the line between divine and non-divine on a logical basis. If the νόθοι καὶ παρέγγραπτοι are included among the gods, then everything and everyone will turn out to be a god. If, on the other hand, they are excluded from the gods, then all gods can be cancelled. Plutarch was opposed to either result: in the De superstitione he attacks both those who do not recognize gods and those who consider everything divine.Footnote ⁵³ However, he agrees with Cotta that theological reasoning is incapable of putting a stop to the sorites. For Plutarch the solution is simple: if theological reasoning is not required to underpin religion, then the sorites, while remaining logically valid, loses its devastating effect. The νόθοι καὶ παρέγγραπτοι can simply be excluded without harming the category of legitimate gods such as Eros. The same goes for deified πάθη, an adjacent or even assimilated category to which the sorites was applied. This immunization strategy is compatible with the stance of Cotta, whose isolated endorsements of theological scepticism and traditional religion may seem puzzling otherwise. In sum, Cotta’s and Plutarch’s uses of the νόθοι καὶ παρέγγραπτοι in contexts that recall theological sorites arguments are two sides of the same coin (a coin that becomes more polished if one accepts the emendations to Cic. Nat. D. 3.39 and Plut. Amat. 756C I have proposed). Cotta invokes the illegitimate gods to point out the impossibility of excluding them in rational theological discourse. Plutarch, on the other hand, mentions them to show how traditional religious discourse restores the possibility of excluding them.

Lucian’s Concilium deorum also uses theological sorites-style arguments and an appeal to traditional religion. Its concern with νόθοι καὶ παρέγγραπτοι does not, as Oliver argued, depend on the actuality of Lucian’s day, but evokes an argumentative tradition that can be traced back to Carneades.Footnote ⁵⁴ As opposed to Cicero and Plutarch, Lucian nevertheless warns us, through the self-defeating character of Momus’ decree, that we should regard with suspicion the simple solution of checking the sorites by applying traditional religion. And indeed Cotta’s mos maiorum and Plutarch’s πάτριος πίστις, the instruments used to stop the sorites, are hardly undisputed static bodies of knowledge. For Plutarch, πάτριος πίστις meant putting Eros (in the Amatorius), the Egyptian Osiris (in the De Iside et Osiride) or Apollo (in the De E) on the highest pedestal: few traditionalists would have agreed with that and some may even have found that Plutarch, not unlike Momus, was disobeying his own rules by pushing a πάθος (Eros), a foreign deity (Osiris) or an Olympian whom tradition equivocally considered Zeus’s subordinate (Apollo). Cotta’s mos maiorum can be approached in a similar way, albeit much more speculatively, given Cotta’s refusal to voice positive opinions. Like Plutarch, he seems to want to exclude the nothi et adscripticii—Heracles, Asclepius, Castor and Pollux, and Romulus, the recognition of whose divinity he associates with imperiti and indocti (Nat. D. 3.39–40)—on the basis of traditional religion. Among his sources for the mos maiorum are Romulus and the Sybilline Books (Nat. D. 3.5). However, the former is said to have been responsible for the recognition of Heracles as a god (Varro fr. 35 Cardauns), while the latter source sparked many third-century innovations in Roman religion, such as the introduction of Asclepius as a god (Livy 10.47.6–7).Footnote ⁵⁵ Cotta’s mos maiorum, which is supposed to be able to exclude the nothi et adscripticii, turns out to diverge at least in some aspects from the sources that he claims to follow.

By invoking tradition in order to demarcate gods from non-gods, Plutarch and Cotta try to save their pantheon from the claws of the sorites logic that leads to a flood of new gods—a flood that Stoic theology, as they understood it, could not evade.Footnote ⁵⁶ At the same time, they present us with a pantheon that is perhaps more their own than their forefathers’. Their apparently conservative resistance to religious innovation, then, turns out to have innovative potential.Footnote ⁵⁷ In Plutarch’s case, at least, it is safe to say that what he calls πάτριος πίστις is a purification of traditional religion. This purification is executed on the basis of a rational (in this case, Platonic) theology: Plutarch’s highest ‘traditional’ gods bear a striking resemblance to the demiurge from Plato’s Timaeus.Footnote ⁵⁸ In this regard, the separation of traditional religion (involving πάτριος πίστις) and rational theology (involving λόγοι) turns out to be less strict than advertised.Footnote ⁵⁹ Once again, Cotta gives us less to go on. There is, however, another Academic sceptic among the characters in the De natura deorum: the character Cicero, whose stated preference for the Stoic Balbus over his fellow Academic Cotta at the end of the dialogue (Nat. D. 3.95) has sparked much debate. R. Woolf argues that the character Cicero is depicted as surpassing Cotta by being more open to positive (in this case, Stoic) positions and thus attempting to bridge the gap between tradition and argument.Footnote ⁶⁰ The author Cicero seems to concur when, in the introduction to the dialogue, he insists on the connection between cognitio and religio (1.1).Footnote ⁶¹ This may signal Cicero’s suspicion that the sceptical separation of traditional religion and rational theology becomes problematic as soon as positive opinions about the nature of the gods are ventured.

The immunization strategy used by Plutarch and Cotta may thus turn out to be untenable, as any effort to purify tradition involves argument. And Lucian? He is mocking all parties involved, as per usual. Through Momus’ sorites-style accusations, those who (like the Stoics) claim to be able to build religion on rational theology are teased. Through Momus’ self-defeat, however, those who (like the sceptical Academics Cotta and Plutarch) criticize the previous group and claim that traditional religion can be the single basis for engaging with the divine come in for playful criticism as well. As always, it is Lucian who gets the last laugh.