The Epicurean Colotes, in a work entitled Περὶ τοῦ ὅτι κατὰ τὰ τῶν ἄλλων φιλοσόφων οὐδὲ ζῆν ἔστιν,Footnote 1 cited two fragments of Empedocles in order to prove that the poet denied that existence exists (frr. B 8–9 DK = 12–13 Wright). Both are prominent in controversies about Empedocles’ physics and his usage of the term φύσις,Footnote 2 but fr. 9 is very corrupt. To have any hope of restoring it, we will need to examine carefully Plutarch's explication de texte in his Adversus Colotem (10–12, 1111f–1113e). Although there have been two detailed treatments of this passage, neither has resolved the difficulties.Footnote 3
In the first fragment, fr. B 8, Empedocles argues that there is no coming to be or passing away, but only the mixture and dissolution of the elements. According to Aëtius, this discussion of Empedoclean μεῖξιςFootnote 4 was in his On Nature Book 1.Footnote 5 Since Aëtius’ readings are nearly all superior to Colotes’ (most editors of Empedocles rightly adopt them),Footnote 6 the Epicurean probably relied on a source who quoted from memory, as did Aristotle.Footnote 7 Here, however, is the text as Colotes quoted it, with line 5 added by me:
One of Plutarch's complaints against Colotes is that the Epicurean excerpts his sources inaccurately or incompletely.Footnote 9 In this case Colotes omitted the concluding reference to death that I have supplied, in order to argue that Empedocles denied the existence of ‘existence’ itself, which, he claimed, the poet here calls φύσις. Plutarch rightly objects that Empedocles actually uses φύσις to mean ‘generation’ or ‘birth’ here; as the latter observes, evidently drawing on knowledge of the original poem, the fact that the poet opposes φύσις to θάνατος proves that this is the sense.Footnote 10
Plutarch goes on to charge that, if the Epicureans equate mixture with generation and dissolution with destruction, they themselves do away with existence; his long tirade makes clear that he was motivated by hostility to what he perceives as atheism on their part.Footnote 11 He then accuses Colotes of being unable to interpret language, rightly in this case.Footnote 12 Plutarch notes that Epicurus himself used the periphrasis ἡ τῶν ὄντων φύσις for τὰ ὄντα.Footnote 13 He puts words into Epicurus’ mouth, making him grant that he employs ordinary language by convention, and end by quoting to that effect Empedocles fr. 9.4, νόμῳ δ’ ἐπίφημι καὶ αὐτός:
κἂν εἴ τις ἔροιτο ‘τί λέγεις, ὦ Ἐπίκουρε, τὸ μέν τι κενὸν εἶναι, τὸ δὲ φύσιν κενοῦ;’, ‘μὰ Δία,’ φήσει, ‘νενόμισται δέ πως ἡ τοιαύτη τῶν ὀνομάτων ὁμιλία· “νόμῳ δ’ ἐπίφημι καὶ αὐτός”’.
If someone were to object ‘Epicurus, why do you say that “void” exists, and “the nature of void” exists?’ ‘Βut by Zeus,’ he will say, ‘such usage of words is standard: “I too speak by convention myself”.’ (Adv. Col. 11, 1112f)
Next Plutarch argues that Empedocles too employed the standard usages of φύσις and θάνατος, even though the poet criticized such usages; to prove his point, he cites the heavily corrupted fr. 9. Ι reproduce Pohlenz's Teubner text as revised by Westmann, and my translation tries to replicate the verses’ obscurity:Footnote 14
τοσοῦτον ⟨δ’⟩ (add. Pohlenz) ἐδέησε τοῦ κινεῖν τὰ ὄντα …, ὥστε μηδὲ τὴν φωνὴν ἐκβαλεῖν ἐκ τῆς συνηθείας, ἀλλ’ ὅσον εἰς τὰ πράγματα βλάπτουσαν ἀπάτην παρεῖχεν ἀφελὼν αὖθις ἀποδοῦναι τοῖς ὀνόμασι τὸ νενομισμένον ἐν τούτοις·
⟨ἃ⟩ (add. Meziriacus) ὁ Κωλώτης παραθέμενος οὐ συνεῖδεν ὅτι φῶτας μὲν καὶ θῆρας καὶ θάμνους καὶ οἰωνοὺς ὁ Ἐμπεδοκλῆς οὐκ ἀνῄρηκεν, ἅ γέ φησι μ⟨ε⟩ιγνυμένων (correxi) τῶν στοιχείων ἀποτελεῖσθαι, τοὺς δὲ τῇ συγκρίσει ταύτῃ καὶ διακρίσει ‘φύσιν’ τινὰ καὶ ‘πότμον δυσδαίμονα’ καὶ ‘θάνατον ἀλοίτην’ (corr. J.G. Schneider: ἀλοιτὴν codd.) ἐπικατηγοροῦντας ᾗ σφάλλονται διδάξας οὐκ ἀφείλετο τὸ χρῆσθαι ταῖς εἰθισμέναις φωναῖς περὶ αὐτῶν. ἐμοὶ μέντοι δοκεῖ μὴ τοῦτο κινεῖν τὸ ἐκφορικὸν ὁ Ἐμπεδοκλῆς ἀλλά, ὡς πρότερον εἴρηται, πραγματικῶς διαφέρεσθαι περὶ τῆς ἐξ οὐκ ὄντων γενέσεως, ἣν ‘φύσιν’ τινὲς καλοῦσι.
Empedocles was so far from disturbing reality … that he did not expel the word (sc. physis)Footnote 16 from his usage, but, abolishing only the harmful misunderstanding of reality that it produces, he restored to the words (i.e. physis and thanatos) their conventional meaning in the following verses:
In citing <these verses>, Colotes was unaware that Empedocles does not do away with people, animals, plants and birds, which, he says, arise when the elements are mixed. Rather, he teaches those who label this compounding and separation ‘birth’ (physis), ‘unhappy fate’ and ‘avenging death’ that they are mistaken, yet he did not abolish the use of the usual words for them. Empedocles seems to me not to alter this mode of expression but, as was said earlier (sc. 1113a), to disagree on a point of fact about generation from non-existent things, which some call ‘birth’ (physis). (Adv. Col. 11, 1113a–12, 1113c)
As transmitted, Empedocles’ verses cannot prove Plutarch's point that the poet also used φύσις in its ordinary sense of ‘nature’, because they do not contain the word φύσις at all.Footnote 17 But Plutarch's last paragraph certainly proves that he thought Empedocles used the words φύσιν, πότμον δυσδαίμονα and θάνατον ἀλοίτην in the verses that he cited. δυσδαίμονα πότμον duly appears in verse 3. θάνατον ἀλοίτην does not appear, and therefore received a separate fragment-number in Diels–Kranz (fr. B 10). However, it is all too likely, since the Adversus Colotem is badly transmitted via two late medieval codices only, that a verse containing these words has fallen out after verse 4; if it began with η like the next line, the loss of a verse by homoearchon is easy to posit.Footnote 18 But where is φύσις, which, if Plutarch is right (and there is no reason to distrust him), Empedocles must have used here in its standard sense of ‘nature’ rather than of ‘generation’?
There have been dozens of efforts to repair verse 1; most involve taking μιγέν ‘the thing mixed’ as the subject and aither as the place where the new life-form appears.Footnote 19 None can be called satisfactory. To my knowledge, nobody has proposed that φύσις appeared in this fragment, but I hope to have shown that such a reconstruction is inevitable. Its rightful place is precisely in verse 1; the order in which Empedocles uses the words φύσιν, πότμον δυσδαίμονα and θάνατον ἀλοίτην will then match the order in which Plutarch mentions them. In later Byzantine script, φύσις could readily have been confused with φῶς. Moreover, immediately before φῶς ‘light’ Empedocles says φῶτα ‘wight’, where I have used the English archaism to render φώς ‘man’, as unfamiliar and obscure to later Greeks as ‘wight’ now is to us. As Diels already saw,Footnote 20 φώς and/or φῶς could well have appeared in a supralinear Byzantine gloss, meant to explain that φῶτα is not from φῶς. Greeks were well aware of the homonymy, since Parmenides wittily adapts Homer's phrase ἀλλότριος φώς ‘alien wight’ (Od. 19.219) as ἀλλότριον φῶς to describe the moon's ‘alien light’ (fr. B 14), and Empedocles borrowed the latter phrase for the moon.Footnote 21 However, the poet would have been extraordinarily clumsy to put φῶτα next to φῶς, a quasi-homonym with a completely different sense. Hence I propose that φῶς is an error for φύσις.
If this is so, verse 1 meant ‘when the nature of <***> was mixed with aither in a man’, referring to the mixture of elements, as Plutarch makes clear (ἅ γέ φησι μ⟨ε⟩ιγνυμένων τῶν στοιχείων ἀποτελεῖσθαι, 1113b); Empedocles used the same periphrasis as that in Epicurus fr. 76, i.e. ‘the φύσις of p’ in place of p. Thus he specified two elements which are mixed in living beings, although the mixture would have varied both according to their speciesFootnote 22 and in their different parts.Footnote 23 What is their main constituent, which must have been specified in the lost disyllable at the end of verse 1?
Since aither or fire is already named, it must be air, water or earth. It can hardly be air or water, since the living creatures that Empedocles lists are more solid than these substances. Hence the missing term must be earth, which would clearly be present in all these creatures. We have a few actual recipes for the mixture in different animals or parts of animals. Empedocles called the body surrounding the soul ἀμφιβρότην χθόνα ‘earth that surrounds a mortal’.Footnote 24 According to him, bone consists of four parts of fire, two of earth and two of water;Footnote 25 the first humans contain water and aither as they sprout from the ground (and the latter origin also implies that they contain earth);Footnote 26 blood and flesh have equal parts of all four elements;Footnote 27 shellfish have the earth on the outside, and stags’ horns are made of earth;Footnote 28 and plants concentrate earth in their roots but aither in their shoots.Footnote 29 Since this passage goes on to list all the kinds of life, both animal and vegetable, that result from this mixture, the elements which they hold in common should be water, earth and fire, which supplies vitality and motion (Empedocles is unlikely to have understood that plants respire). But there are only two of them here; given the choices, earth is surely preferable, since mixtures of fire and air or fire and water do not produce solid bodies. Thus the missing disyllable can easily be restored as γαίης: Empedocles will then be using φύσις … γαίης as a periphrasis for γαῖα. The aorist passive neuter singular participle μιγὲν needs to be a singular finite verb governed by φύσις: the third person singular μίγη lies ready to hand.Footnote 30
Thus fr. 9 of Empedocles should be restored and translated as follows:
