Navigation – Plan du site
Dossier : Les Formes platoniciennes dans l'Antiquité tardive

Making a Virtue of Necessity : Anangkê in Plato and Plotinus

Peter Adamson
p. 9-30

Texte intégral

  • 1 This is taking at face value On Providence §22 ; tellingly Proclus emphasizes the “mortal” nature o (...)

1According to Neoplatonists, the world is in some sense perfect. As Proclus says in his commentary on the Timaeus, it is a “whole of wholes”, lacking in nothing. Of course, the physical world is perfect only according to some standard lower than that of the intelligible world ; still, it attains a perfection appropriate to it. On the other hand, particular things within the physical world are often far from perfect. Neoplatonists had their own intimate experience of this fact : Plotinus had terrible eyesight, Porphyry was subject to bouts of suicidal melancholy, and apparently Proclus’ house burnt down.1 Physical things are caused exclusively by intelligible things, which are perfect, and yet individual physical things fail to duplicate this perfection. This calls out for an explanation. An obvious thing for any Platonist to say is that we have been given the explanation in the Timaeus :

“This ordered world is of mixed birth : it is the offspring of a union of necessity and intellect. Intellect prevailed (ἄρχοντος) over necessity by persuading (πείθειν) it to direct most of the things that come to be toward what is best (τῶν γιγνοµένων τὰ πλεῖστα ἐπὶ τὸ βέλτιστον ἄγειν), and the result of this subjugation of necessity to wise persuasion (ὑπὸ πειθοῦς ἔµφρονος) was the initial formation of this universe.” (Tim. 48a1-5 ; Zeyl trans.)

2But what does it mean to say that in making the world, intellect had to contend with, and ultimately overcome, “necessity” ? Part of the answer is easy to give : things may not be perfect, but they are as perfect as possible. In other words, the reason that physical things are not perfect is that it is impossible for them to be any better than they are. Or, in still other words, it is “of necessity” that they are imperfect.

3Of course this will only be a good answer if we can further say where the “necessity” comes from, and what sort of necessity is involved. It doesn’t look like a straightforward logical necessity, in the sense that no contradiction seems to be entailed by the presence of perfection, or to put it conversely, by the absence of evil. (Indeed, it has often been argued that the presence of evil is logically incompatible with the existence of purely good divine principles.) Instead it seems likely, especially in the context of Neoplatonism, that the necessity is a metaphysical one. That is, there are metaphysical reasons why it is impossible for intelligible Forms to be perfectly replicated by the physical things that participate in them. Perhaps this follows immediately from the sheer fact that these things are copies or images : a truly perfect “copy” or “image” is in fact not a copy or image at all, but would be indistinguishable from, or even identical to, the original. Alternatively one might invoke the nature of physical things themselves, arguing for instance that spatial extension brings with it the possibility of imperfection. For instance, spatially distinct things can interfere with each other, and will thus inevitably impair one another’s proper functioning. Again, in the context of Plotinus’ theory of evil, one might point out that the existence of matter – which is the principle of evil – is itself metaphysically necessary.

  • 2 For a subtle discussion of some of the same themes in a related context, see A. Silverman, “Ascent (...)
  • 3 G. Van Riel, “Proclus on Matter and Physical Necessity”, in R. Chiaradonna and F. Trabattoni (eds), (...)

4Such accounts in terms of metaphysical necessity may go some way towards explaining what “necessity” means in the Timaeus. But I think they fall short of capturing the full sense in which imperfections are “necessary” for Plato and for later Platonists. This can already be seen by considering the passage from the Timaeus quoted above. If something’s being “necessary” simply means that it “cannot be otherwise”, it is hard to understand what it might mean to “persuade” necessity, or to say that necessity is one of the progenitors of the physical world. A closer consideration suggests that the notion of anangkê combines several simpler ideas into one. That which “cannot be otherwise” is, at most, only one component of this complex notion. In this paper, then, I want to explore what else anangkê might mean in Plato. My interpretation is intended to give a reasonable reading of Plato himself, but it will be clear that I am reading him in a way that Neoplatonists would find congenial. For instance, all my focus will be on necessity as an explanation of physical imperfection (as opposed to other features of physical things that are “unavoidable” but unobjectionable).2 I will then move on to the Neoplatonists themselves, discussing how they used Plato’s nuanced sense (or rather senses) of anangkê in their attempt to explain the fact that physical things fail to replicate intelligibles perfectly. For reasons of space, and because another recent paper has explored the meaning of anangkê in Proclus,3 I will concentrate on Plotinus with a brief concluding glance at how Proclus reacted to the Plotinian understanding of anangkê.

Plato : the Timaeus and the Republic

  • 4 One reason for the caveat is that the skull, and everything else fashioned by demiurgic activity, w (...)

5I have already said above that anangkê in the Timaeus seems to mean more than “what cannot be otherwise”. One might go even further and say that it has nothing at all to do with this notion. Consider the well-known reference to anangkê at Timaeus 75a-c, where the thickness of the skull is said to be ἐξ ἀνάγκης (75a7) a compromise : if it were thicker we could live longer, but we would be less intelligent. This passage actually requires that the skull (in some sense4) could have been otherwise. It is just that, given certain goals, it was necessary for it to be as it is. To put this in an Aristotelian way, the features of things produced by divine causes are teleologically directed. This is why, when he first introduces the notion of anangkê into his account, Timaeus says that persuasion “directs things towards the best” (48a, cited above). On the other hand, our admiration for the optimal design of the skull does not eliminate our regret that the skull is as it is. It is a pity that the skull could not be made thicker, for if it had been, we would live longer. At the same time, it is a pity the skull could not be made thinner, for then we would be more intelligent. The “necessity” here thus also refers to a kind of constraint, in this case the constraint that the skull cannot be both thicker and thinner than it is. Again, notice that the skull could have been thicker, and equally, it could have been thinner. So the constraint is not that there are no alternative possibilities, but rather that the desirable possibilities cannot all be realized, because they are incompatible.

  • 5 Silverman, “Ascent and Descent”, p. 61-2 and p. 68, differs from me in thinking that regret implies (...)
  • 6 Though the skull is a particularly striking example of this, it is certainly not the only one we ar (...)
  • 7 Here I would follow T. Johansen, Plato’s Natural Philosophy. A study of the Timaeus-Critias, Cambri (...)

6There are, then, constraints that prevent the demiurge from creating a world which is utterly free from regrettable aspects.5 Plato tries to capture this idea by saying that the demiurge must persuade necessity, and that this persuasion can only go so far. He says, for instance, that the demiurge bestows perfection on elemental bodies “to the extent that the nature of necessity willingly yielded to persuasion” (ὅπῃπερ ἡ τῆς ἀνάγκῆς ἑκοῦσα πεισθεῖσά τε φύσις ὑπεῖκεν) (56c5-6). The famous example of the skull is then supposed to illustrate the results of divine persuasion.6 It is thus an example of what I will call “persuaded necessity”. What is the difference between persuaded and unpersuaded necessity ?7 We have already seen one difference : persuaded necessity is compatible with and perhaps even requires that things could have been otherwise. By contrast unpersuaded necessary is inevitable, and for precisely this reason can explain regrettable aspects of the physical world in a way that deflects blame from the divine causes. We can now see why, at least in the case of persuaded necessity, the notion of anangkê in the Timaeus is a complex and almost paradoxical one. Once it is persuaded, necessity is precisely that which is both teleologically ordered to a goal and regrettable. It is, furthermore, at least compatible with the presence of alternative possibilities. Unpersuaded necessity is by contrast necessary in the more ordinary “modal” sense of what cannot be otherwise ; and it is not (yet) teleologically ordered. The only thing that remains constant is the idea of regret, that is, the idea that what is necessary falls short of the ideal. To summarize, then :

Unpersuaded necessity

Persuaded necessity

(1a) Non-teleological

(1b) Teleological

(2a) Cannot be otherwise

(2b) Can be otherwise

(3) Regrettable

(3) Regrettable

7How unusual is this way of using the word anangkê ? To find out, we can turn to the very useful enumeration of the senses of the word given by Aristotle in Metaphysics Delta :

  • 8 I take this phrase to be a further explication of the first sense of anangkê ; note it begins kai r (...)
  • 9 I have here omitted a fourth sense, which has to do with necessity in demonstration.

“We call the necessary (ἀναγκαῖον) (i) that without which, as a condition, a thing cannot live (οὗ ἄνευ οὐκ ἐνδέχεται ζῆν ὡς συναιτίου), for instance breathing and food are necessary for an animal – for it is incapable of existing without these (ἀδύνατον γὰρ ἄνευ τούτων εἶναι) – and8 the conditions without which good cannot be or come to be (καὶ ὧν ἄνευ τὸ ἀγαθὸν µὴ ἐνδέχεται ἢ εἶναι ἢ γενέσθαι), or without which we cannot get rid or be freed of evil. For instance, drinking the medicine is necessary in order that (ἵνα) we may cured of disease, and sailing to Aegina is necessary in order that we may get our money. (ii) The compulsory and compulsion (ἔτι τὸ βίαιον καὶ ἡ βία), i.e. that which impedes and hinders contrary to impulse and choice (παρὰ τὴν ὁρµὴν καὶ τὴν προαίρεσιν ἐµποδίζον καὶ κωλυτικόν). For the compulsory is called necessary (τὸ γὰρ βίαιον ἀναγκαῖον λέγεται) ; that is why the necessary is painful, as Evenus says : ‘for every necessary thing is ever irksome’. And compulsion is a form of necessity (καὶ ἡ βία ἀνάγκη τις), as Sophocles says : ‘compulsion makes this action a necessity’ (ἀλλ᾿ ἡ βία µε ταῦτ᾿ἀναγκάζει ποιεῖν). And necessity is held to be something that cannot be persuaded (καὶ δοκεῖ ἡ ἀνάγκη ἀµετάπειστόν τι εἶναι) – and rightly, for it is contrary to the movement which accords with choice and with reasoning (ἐναντίον γὰρ τῇ κατὰ τὴν προαίρειν κινήσει καὶ κατὰ τὸν λογισµόν). (iii) We say that that which cannot be otherwise is necessarily so (ἔτι τὸ µὴ ἐνδεχόµενον ἄλλως ἔχειν ἀναγκαῖόν φαµεν ἔχειν). And from this sense of necessary all the others are somehow derived (κατὰ τοῦτο … λέγεταί πως). For as regards the compulsory we say that it is necessary to act or to be acted upon, only when we cannot act according to impulse because of compulsion (κατὰ τὴν ὁρµὴν διὰ τὸ βιαζόµενον) – which implies that necessity is that because of which the thing cannot be otherwise (δι᾿ἣν µὴ ἐνδέχεται ἄλλως). And similarly as regards the conditions (συναιτίων) of life and of good, when in the one case good, in the other life and being, are not possible without certain conditions, these are necessary, and this cause (αἰτία) is a kind of necessity.”9 (Metaphysics V 5, 1015a20-b9 ; Ross trans., adapted)

  • 10 For a comparison of anangkê in the Timaeus to Aristotelian hypothetical necessity, see Johansen, Pl (...)
  • 11 The point would of course also apply to the further sense of the necessary in demonstration : what (...)
  • 12 Similarly, to take Aristotle’s sense (ii), if I am forced to act, I might well say to myself that I (...)

8Aristotle tells us that the most basic sense of anangkê is sense (iii) : the necessary is that which cannot be otherwise. Yet like Plato, he is happy to talk of “the necessary” in situations where there are alternative possibilities. Some things are “necessary” only given the presence of some goal : this is hypothetical or teleological necessity, Aristotle’s sense (i).10 The same is true of his sense (ii), where anangkê is synonymous with “force” (bia). I can be “necessitated” to act by a compelling cause. But if this cause itself might have been absent, then so could my action have been otherwise. Thus, when Aristotle says that his sense (iii) underlies both senses (i) and (ii), he means only that they are instances where something cannot be otherwise under some condition, either a goal as in sense (i), or a compelling cause as in sense (ii).11 In light of this, we can add a caveat to point (2b) : although instances of persuaded necessity in some sense could have been otherwise, they might in another sense be unavoidable. For instance, if the Demiurge inevitably has the power and desire to persuade necessity in the best possible way, then the results will be inevitable. This will be important when we get to Plotinus, since for him too, intelligible causes are inevitably successful in their creative activity. The point of (2b), then, is that things would have been otherwise in the absence of divine ordering. I take it that this thought is a coherent one even if divine order is inevitable.12

  • 13 Evenus might be a better guide in this respect, of course. The use of anangkê in tragedy also shows (...)

9Aristotle’s survey suggests that the way anangkê is used in the Timaeus exploits a range of natural connotations belonging to the Greek word. What is particularly noteworthy is that Aristotle also associates something’s occurring “by necessity” with its being against our will or regrettable. Aristotle does not explicitly indicate that Plato’s sense (3), the regrettable, can come together with sense (1b), the hypothetical. So he does not really give an example of “persuaded necessity”, unless we suppose (not implausibly) that having to drink the medicine or sail to Aegina is a rather annoying thing to have to do. Yet, with regard to sense (ii) he is explicit that anangkê can carry connotations of regret, and in fact he quotes Evenus to this very point : the necessary is “ever irksome”. Of course, taking Metaphysics Delta as a reliable guide to the connotation of Greek words is not without risk.13 We can, however, find further evidence that the Timaeus use of anangkê is a natural one by turning to the Republic’s treatment of desire and pleasure.

10In books VIII-IX, Plato has Socrates introduce the concept in the following passage :

“Aren’t those [desires] we can’t prevent and those whose satisfaction benefits us rightly called necessary ? For it is by nature necessary for us to satisfy them both (οὐκοῦν ἅς τε οὐκ ἂν οἷοί τ᾿ εἶµεν ἀποτρέψαι, δικαίως ἂν ἀναγκαῖαι καλοῖντο, καὶ ὅσαι ἀποτελούµεναι ὠφελοῦσιν ἡµᾶς· τούτων γὰρ ἀµφοτέρων ἐφίεσθαι ἡµῶν τῇ φύσει ἀνάγκη).” (Resp. 558d11-e3 ; Grube/Reeve trans., adapted)

11Socrates is quite emphatic about the use of the term anangkê here. Both here and again at 559a1, he solicits from Adiemantus explicit agreement that it is the right term for such desires. By contrast, desires which one can “get rid of” (ἀπαλλάξειεν) and which do not lead to the good are said to be “unnecessary” (µὴ ἀναγκαίους) (559a3-6) ; later they are called useless or not beneficial (ἀνωφελῶν, 561a3). Only certain desires – and the pleasures associated with their fulfillment – are said to be necessary, in that they contribute to some good, either our survival or a healthy state (εὐεξία, 559b7). These desires and pleasures are, then, a case of the hypothetical sense of anangkê (Plato’s sense (1b), Aristotle’s sense (i)). As we saw above with persuaded necessity, these “necessary desires” are desires whose satisfaction becomes necessary in the light of some goal. For instance, I need to eat in order to avoid starving to death. But they are also regrettable (Plato’s sense (3)), at least to the right-thinking man, who thinks little of necessary pleasures, and “calls them really necessary, since he’d have no need for them if they weren’t necessary” (581e3-4, Grube/Reeve trans.). As for the other desires which we are encouraged to eliminate entirely, they are “unnecessary” precisely because they are not hypothetically ordered towards the maintenance or achievement of some good, such as health.

  • 14 And, interestingly, by means of anangkê itself. The simplest thing to say about anangkê in this pas (...)
  • 15 An interesting parallel in Republic V labels as a matter of necessity the fact that if they dwell t (...)

12We find a further parallel with the Timaeus in that the Republic gives persuasion a role in restraining desires. At the beginning of book IX, Socrates says that some unnecessary desires are in themselves “lawless” (παράνοµοι), but that by means of “laws and better desires along with reason (µετὰ λόγου)”, they can be suppressed and even eliminated (571b5-8), though they will re-emerge during sleep even for the most moderate of men (572b). Though he does not use the term “persuasion” in this passage, earlier in book VIII Socrates has described the oligarchic man as someone who “fulfills only his necessary desires… and enslaves his other desires, since they are vain” (554a5-8). How does he restrain these vain desires ? Not, we are told, by “persuading” (πείθων) the desires and “setting them at rest by means of reasoning (λόγῳ)”, but by means of “compulsion” and “fear” (554d1-3).14 Presumably the virtuous man, by contrast, does use persuasion and reasoning. In both contexts, Socrates speaks only of unnecessary, rather than necessary, desires. Still it is interesting that persuasion, along with the imposition of reason and law, plays a part in the control of desire.15 By these means the virtuous man is able to subdue, and even eliminate entirely, his unnecessary desires, particularly those that are lawless. Furthermore, the virtuous man allows the necessary desires to cause no disruption to virtue. He satisfies them only insofar as they contribute to the goal of keeping him alive and healthy, and he regrets that even this minimal concession to necessary desire is unavoidable.

13Turning back to the Timaeus, we can note that it too recognizes a class of “necessary affections” (παθήµατα), including pleasures and pains, seated in the mortal soul (69c8-d1). The mortal soul itself is said to be necessary (69d5), and is located as far away from the head as possible, so as “not to besmirch the divine soul any more than necessary” (69d7 : ὅτι µὴ πᾶσα ἦν ἀνάγκη). This gives us further warrant for drawing a parallel between the Republic and the Timaeus : necessary desires and pleasures are not just analogous to anangkê in the Timaeus, but are actually named as a specific case of anangkê in the Timaeus. So the case of the virtuous agent is like the case of the demiurge. The virtuous agent must make concessions to his necessary desires, but uses persuasion so as to subdue his desires. Likewise, the demiurge must make concessions to anangkê, but he uses persuasion to subdue it. The result is, in the ethical context, a life of moderation, in the cosmological context, a providentially ordered world. The life of moderation is not ideal, in that it contains the regrettable necessary desires. Nor, by parity of reasoning, are the demiurge’s products ideal ; they are simply as good as they could have been.

  • 16 See G.R. Morrow, “Necessity and Persuasion in Plato’s Timaeus”, Philosophical Review 59, 1950, p. 1 (...)
  • 17 Strange, “Double Explanation...”, p. 28-9.
  • 18 Arguably the presence of spatial extension is already part of unpersuaded necessity (the pre-cosmic (...)

14Does this mean that there are some features of the world that are entirely out of the control of the demiurge ? Presumably not, any more than there are some actions taken by the virtuous man that are out of his control. Rather, although anangkê does serve as a constraint on what the demiurge makes, it is the demiurge and his divine helpers who choose every aspect of the world. Thus I would distance myself slightly from the identification that is often made between anangkê and “mechanical” processes in the Timaeus.16 Steven Strange has distinguished between the “domains” of reason and necessity, saying that the domain of necessity consists in “purely material properties”. As a basic example, he says that it is “not in any way up to the Demiurge that the elements are bodily and spatially extended”.17 And this is certainly right, in that elements must be bodily and spatially extended if they are to have any physical existence at all. Similarly the skull will be relatively fragile if it is to be a fit housing for the human brain. But it is the demiurge who decides that these goals will be achieved in the first place, and there is no hint that anangkê constrains the prior decision of which goals to achieve. If there is any constraint in this respect, it comes not from anangkê but from the demiurge’s own goodness, or from the need for paradigms to be instantiated. Thus even the extendedness of the elements satisfies our criteria (1a) and (1b) : if it is “inevitable” it becomes so only in the light of some goal, without which things would have been otherwise.18

15Having said this, it would seem that in the Timaeus we do find cases of necessity which seem to be non-teleological. I have in mind features of the “pre-cosmic” state in which the Receptacle has not yet received order from the demiurge. Or, if you prefer a non-temporal reading, features of the cosmos as it would be if, counterfactually, it were not being ordered by the demiurge at all times. It is here that we find clear cases of unpersuaded necessity. It is apparently just unavoidable, for instance, that there are traces of the elements in the receptacle (53b) or that the receptacle has a shaking motion (53a-b). And the traces and shaking motion are not teleologically ordered towards any good. Actually, now that we have some concrete cases to consider, we might begin to wonder whether it is really true that unpersuaded necessity is regrettable. Does it really make sense to regret the presence of traces in the receptacle, or regret the shaking motion of the receptacle ? In fact, if such things are really wholly inevitable, isn’t it rather irrational to regret them ? It would be like regretting that three plus two equals five. The answer, it seems to me, is that we do not regret them directly or as such. Rather, we regret the pre-cosmic or non-ordered aspects of the world precisely insofar as they are the ultimate sources of constraint on the demiurge.

  • 19 Again the Republic might leap to mind, at least for a reader (like a Neoplatonist) who assumes that (...)

16The identification of these pre-cosmic or non-ordered aspects as cases of unpersuaded necessity might also make us wonder what sense it makes to talk about “persuading” them. Perhaps Plato does so because the demiurge is like a rhetorician whose logoi must be tailored to his audience. It is the rhetorician who decides what they will believe, with certain arguments or blandishments being necessary to his speech only insofar as they will help him instill the desired beliefs in the audience. Again these rhetorical devices are hypothetically necessary, and they might in some cases also be regrettable even though they are optimally fit for purpose – the noble lie of the Republic leaps to mind. Notice that it is indeed “beliefs” that are at issue here : the Timaeus itself tells us that belief, and not nous, is the product of persuasion (51e). Instead, as we saw in the passage from which we began (Timaeus 48a), it is nous that does the persuading, and necessity is its “audience”. This invites us to draw a parallel between the epistemic state of belief, and the ontological state of the world of becoming made by the demiurge. For both are brought about by persuasion.19

  • 20 This parallel is also drawn by Van Riel in “Proclus on Matter and Physical Necessity”.

17All of this underscores the point that the demiurge decides which goals will be achieved, just as the rhetorician decides which beliefs will be adopted by the audience. Just as the nature of the audience determines the way persuasion will come about, so (unpersuaded) necessity places constraints on how the demiurge’s intentions will be fulfilled.20 One might object to this that there are certain beliefs that simply cannot be instilled in a given audience. In this sense, it is not entirely up to the rhetorician which goals to choose, because some goals cannot be realized at all. In the demiurgic case, the parallel objection would be that there are some things the demiurge simply cannot make, some goals he could not have fulfilled even if he wanted to. This objection fails on both sides. Rhetoricians like Gorgias did at least claim the ability to instill beliefs in an audience on any subject (Gorgias, 456a-457c ; admittedly, it is not quite clear whether this implies the ability to instill any belief regarding any subject, but I suspect Gorgias would say that it does). And more importantly for us, the demiurge is able to bring about a complete image of the Forms (or the paradigm Living Being) in the cosmos ; nothing is left out. Anangkê concerns only how the paradigm is instantiated, not the extent to which it is completely or incompletely instantiated.

  • 21 Again I agree here with Johansen, who criticizes Grote’s characterization of anangkê as “force… wit (...)

18What we have in the Timaeus, then, is an optimistic cosmology according to which no necessity remains unpersuaded. The cosmos is entirely suffused with form and order through demiurgic activity. To see what the anangkê of the Timaeus therefore is not, we could recall Aristotle’s sense (ii) of necessity, namely force or compulsion. He says that this sort of necessity “is held to be something that cannot be persuaded” (ἀµετάπειστον ; compare µεταπειστόν at Timaeus 51e4). Plato could agree with Aristotle that when anangkê means bia, then anangkê cannot be persuaded. It is just that anangkê does not mean this in the Timaeus.21 Rather, Plato holds that all the cases of necessity we actually encounter are cases that have, by persuasion, been “directed towards the best”. Unpersuaded necessity is present in the Timaeus only as a postulated, underlying cause which accounts for the fact that these cases of persuaded necessity often have regrettable aspects.

Plato : the Theaetetus

19But Plato is not always this optimistic. His most celebrated passage on theodicy is not in the Timaeus but the Theaetetus. In this passage, Theodorus says that there would be less evil in the world if everyone were persuaded (he uses the verb πείθω, 176a3) by Socrates’ previous speech. Socrates replies :

“It is not possible that evil should be destroyed, Theodorus – for there must always be something opposed to the good (ἀλλ᾿ οὔτ᾿ ἀπολέσθαι τὰ κακὰ δυνατόν, ὦ Θεόδωρε, ὑπεναντίον γάρ τι τῷ ἀγαθῷ ἀεὶ εἶναι ἀνάγκη) – or that it should have its seat in heaven (ἐν θεοῖς). But it must necessarily (ἐξ ἀνάγκης) haunt human life, and prowl about this earth.” (Theaetetus 176a5-8 ; Levett/Burnyeat trans., adapted)

20This passage seems to use a different notion of necessity than the one we have seen in the Timaeus. Here the two references to anangkê look like they simply restate the impossibility of eliminating evils (οὔτ᾿ ἀπολέσθαι τὰ κακὰ δυνατόν), so that we have a more basic, “modal” sense of necessity : Aristotle’s sense (iii), that which cannot be otherwise. Of course this does not, by itself, suggest any tension or conflict with the Timaeus. After all, as we have seen unpersuaded necessity in the Timaeus is also necessary in this sense (this was feature (2a)). And besides, it isn’t clear that the unavoidable and sometimes regrettable aspects of the physical world, which are in focus in the Timaeus, are to be identified with “the evils” (τὰ κακά) of this passage.

21Still, it is not hard to see a potential conflict between the two texts. It’s a familiar point that the Greek kakon is much broader than the English word “evil”, which has a strong connotation of moral evil in particular. If the Theaetetus is trying to explain the presence of badness in the world, then it is more tempting to try to bring it into line with the Timaeus account of imperfection. Against this, one might say that the previous conversation in the Theaetetus suggests a focus on moral evils. Yet Plotinus will refer to this passage in contexts where deficiencies, “evils” in the broad sense, are at stake. For him, at least, this passage from the Theaetetus and the Timaeus account of necessity are intended to explain the same phenomenon, and as we will see he duly quotes the two texts together in his discussions of evil. This means that Plotinus must somehow reconcile the Timaeus account with the Theaetetus’ remark that evils are ineliminable because “there must always be something opposed to the good”. This looks like a reference to necessity that is not only unpersuaded, but that inevitably remains unpersuaded. If it was right to say that in the Timaeus the demiurge is able to subject all necessity to persuasion, then we seem to have a straightforward contradiction between the two texts.

22This tension is obscured, to some extent, by the fact that in both dialogues the existence of the bad (to use a neutral term) is explained by appealing to anangkê. The superficial agreement masks a deeper disagreement, since in the Timaeus, this term refers to something that may well be (3) regrettable, but is nonetheless (2b) chosen from among other possibilities and chosen precisely because (1b) it is the possibility most conducive to the good. By contrast, in the Theaetetus the bad is present because it cannot be eliminated, and is regrettable because it is actually the opposite of the good. So evil in the Theaetetus is not to be identified with persuaded necessity. And in fact, evil in the Theaetetus is even more regrettable than unpersuaded necessity in the Timaeus. The latter was (1a) non-teleological, and (3) regrettable insofar as it is ultimately responsible for the compromises the demiurge must make in shaping our world. But at least it does submit to the persuasion of the demiurge. By contrast evil in the Theaetetus is not just non-teleological, but, as it were, anti-teleological, insofar as it is the opposite or contrary, rather than the mere privation, of the good. So to summarize again :

Unpersuaded necessity

Persuaded necessity

Evil in the Theaetetus

(1a) Non-teleological

(1b) Teleological

(1c) Opposite of the good

(2a) Cannot be otherwise

(2b) Can be otherwise

(2c) Cannot be otherwise

(3) Regrettable

(3) Regrettable

(3) Regrettable

23The difference between (1a) and (1c) is, again, that the necessity of the Timaeus is wholly susceptible to persuasion by the good, whereas evil in the Theaetetus remains opposed to the good.

Plotinus

24It is unsurprising that the word anangkê has a similar range of notions for Plotinus. He was in general a careful reader of Plato, and his remarks on the topic of anangkê in particular engage with the very texts we have been examining in the Timaeus and Theaetetus. These appear especially in the context of his treatments of evil. It is a familiar, if not over-familiar, point that Plotinus’ views on evil sometimes seem relatively optimistic, as when he is engaged in polemic against the Gnostics, and sometimes relatively pessimistic, as when he identifies matter itself as the principle of evil. So there is nothing new in saying that there is a tension in Plotinus between optimism and pessimism regarding evil (or more broadly “badness”, i.e. whatever is regrettable). But in this section, I will make a case for the narrower and less familiar point that Plotinus uses all three of the senses of “necessity” just enumerated in his treatment of evil.

  • 22 On this treatise see A. Schniewind, L’Éthique du sage chez Plotin : le paradigme du spoudaios, “His (...)

25For Plotinus as for Plato, it will be useful to see how the word anangkê is used in ethical, as well as cosmological, contexts. This will reveal that the connotations of the word for Plotinus are closely related to the ones we have found in Plato. One relevant text is Enneads I.4 [46], which describes the needs of the body as “necessities” from the viewpoint of the virtuous man or sage (σπουδαῖος).22 Like Plato’s virtuous man, Plotinus’ sage does pursue physical things, but only because he must :

  • 23 For a similar use of anangkê see II.9 [33] 9.11.

“These things, which have no attraction for us when they are there and do not contribute anything to our happiness, but which we seek in their absence because of the presence of things which distress us, can reasonably be called necessities, but not goods (ἀναγκαῖα ἀλλ᾿ οὐκ ἀγαθά).”23 (I.4 [46] 6.27-30 ; Armstrong trans.)

26Plotinus tends to be even more dismissive of these necessities than Plato was in the Republic : he says they make no contribution to our overall telos, namely happiness, but contribute merely to our continued existence (πρὸς τὸ εἶναι, I.4 [46] 7.2-3). No mention is made even of the Republic’s idea of a “healthy state” (εὐεξία). Here the emphasis is mostly on the necessary being regrettable, because the goal towards which the necessities are ordered is a derisory one – the avoidance of distracting pain, for instance. Nonetheless, they remain purposive.

  • 24 A related use of anangkê can be found in Plotinus’ treatment of fate (heimarmenê), which he describ (...)
  • 25 The cases may not be exactly analogous. For one thing the sage is really the agent of virtuous deed (...)
  • 26 As in I.4, Plotinus stresses that although these teloi are accepted by the sage, they are not to be (...)

27In a well-known passage from Enneads VI.8 [39], Plotinus extends the idea of anangkê to virtuous action. He remarks that practical virtuous action is a response to some “necessity” which has arisen but which one regrets. For instance the courageous man regrets the war in which he fights, just as Hippocrates would wish that no one was ever ill. In general, “virtue is always necessitated (ἀναγκαζοµένης) to perform one deed or another” (VI.8 [39] 5.11-13). Plotinus holds that the element of necessitation here prevents virtuous action (praxis) from being “purely” voluntary or “up to” virtue itself (VI.8 [39] 5.22). But neither, of course, is the courageous man’s standing in battle wholly involuntary, as it would be if actual compulsion were brought to bear on him (cf. the mention of bia at VI.8 [39] 1.33).24 Rather, I suggest that the virtuous man feels much the same way about the occasions of practical virtue as the sage feels about his bodily needs.25 These things present a certain “necessity” in that one would wish for their absence, but given that they are present, one ought to deal them in a fashion consistent with virtue. Things like eating and standing fast in battle are “necessary” with respect to some telos accepted by the sage,26 for instance the avoidance of bodily disturbance or the instantiation of virtue.

  • 27 For this treatise see D. O’Meara, Plotin. Traité 51, “Les écrits de Plotin”, Les Éditions du Cerf, (...)
  • 28 The literature on evil in Plotinus is of course vast and contentious. A fundamental study is D. O’B (...)

28So these passages show that, at least in ethical contexts, Plotinus does operate with a notion of anangkê that combines goal-directedness with the idea of regret. Does he also do so in the context of explaining imperfections in the physical world ? Here we can do no better than to turn to Enneads I.8 [51], one of Plotinus’ last works and his most sophisticated discussion of the cosmological problem of evil.27 This is the problem or set of problems set out at the beginning of the treatise : what is evil, what is its nature, where does it reside, and is it one of the things that exist (I.8 [51] 1.3-6) ?28 Famously, Plotinus answers that evil is matter, even though matter derives ultimately from the Good. Matter is a source and condition of all deficiency and evil, even for the soul. We can gratefully sidestep most of the controversies surrounding the nature and origin of evil, concentrating only on the question of why and in what sense matter or evil is necessary. At first glance, what Plotinus says about this in I.8 [51] looks to be a variation on the rather unilluminating claim that evil exists because there could not have been no evil. If this were right, the meaning of anangkê would be restricted to Aristotle’s third sense (“cannot be otherwise”), and Plotinus’ efforts would be devoted to explaining why evil is anangkaion in this sense. As we saw, one basis for such a view would be Theaetetus 176-7, which he quotes in the following passage :

  • 29 Note that Plotinus replaces οὔτ᾿… δυνατόν with simply µή, one of several variations from the origin (...)

“But one should investigate in what sense [Plato] says that ‘evils’ are not29 ‘eliminated’ but exist of ‘necessity’… For [Plato] evils are vice (κακία) and those things deriving from vice. When the interlocutor says that evils would be removed (ἀναίρεσιν) ‘if [Socrates] persuaded men of what he says’ (εἰ πείθοι τοὺς ἀνθρώπους ἃ λέγει), [Socrates] says, ‘this is impossible (µὴ δύνασθαι) ; for there are evils by necessity (ἀνάγκῃ), since there must be (δεῖ εἶναι) the contrary of the good’.” (Enn. I.8 [51] 6.1-17)

  • 30 For Plotinus’ argument to this effect see D.J. O’Meara, “The Metaphysics of Evil in Plotinus : Prob (...)

29In what follows Plotinus considers the question of whether there is really any “absolute necessity” (ἀνάγκη πανταχοῦ) that if the good exists, evil must exist as well. He concludes that although this is not necessary for every pair of contraries, it does hold in the case of good and evil.30 Here everything points towards the idea that “necessarily” means nothing more than “cannot be otherwise”. The same is true of the conclusion of the following chapter, which argues that there is “necessarily” a last thing to come from the good, and this “last thing” (τὸ ἔσχατον) will retain no good at all (I.8 [51] 7.16-23).

30But between chapter 6 and the end of chapter 7, we have the following passage :

“But how then is it necessary (ἐξ ἀνάγκης) that if the good exists, so should evil ? Is it because there must (δεῖ) be matter in the universe ? For this universe is necessarily (ἐξ ἀνάγκης) composed of contrary principles ; it would not exist at all if matter did not exist. ‘For the’ nature ‘of this universe was a mixed result of the combination of intellect and necessity’.” (I.8 [51] 7.1-5 ; Armstrong trans., adapted)

  • 31 By contrast to the Theaetetus quotation this is fairly accurate, with the notable exception of subs (...)
  • 32 L. Gerson, Plotinus, Routlege, London, 1994, p. 192-7, also invokes hypothetical necessity in this (...)

31Here we have a quote of our opening Timaeus passage (47e5-48a1),31 and, accordingly, a gesture in the direction of a more optimistic understanding of evil. The existence of matter is not here explained merely by invoking the necessary truth that the good must have its contrary. Rather, Plotinus points out that there must be matter if this universe is to exist. This suggests that matter is after all hypothetically necessary, despite being the source and even nature of evil. In light of this we might also rethink Plotinus’ claim that evil must exist as the contrary of good. Would it be fair to understand this as meaning that, if there is to be good, there must also be evil ?32 That would associate matter and evil with the hypothetical nature of persuaded necessity in the Timaeus.

32Certainly elsewhere, Plotinus does hint that evil or deficiency might be hypothetically necessary to some telos, with the telos being either the good or the existence of the physical universe. For instance :

“Are the evils in the universe necessary (ἀναγκαῖα), because they follow things that precede them ? Or rather because, if there were not evils, the universe would be incomplete (ἀτελές).” (Enn. II.3 [52] 18.1-3)

33We also get hints of this idea in the great treatise on providence (III.2-3 [47-8]), in one case quoting the same passage from the Theaetetus :

  • 33 This is a rather difficult passage, situated as it is in the midst of a dialectical consideration o (...)

“That argument is no longer any help… which maintains that all is as well as it can be, and that it is the fault of the matter when things are so disposed as to be less than good, and that ‘evils cannot be done away with’. If, that is, it is really true that things had to be so, and that it is well that they should be so, and matter does not come along and dominate (κρατεῖ) but was brought along so that (ἵνα) things should be in this state, or rather is itself, too, caused to be as it is by the rational principle. The rational principle, then, is the origin, and all things are reason (πάντα λόγος).”33 (Enn. III.2 [47] 15.7-14 ; Armstrong trans.)

  • 34 Though it is common in the secondary literature to ascribe this claim to Plotinus, as far as I can (...)
  • 35 James Wilberding has suggested to me a compelling argument against identifying matter with persuade (...)
  • 36 The idea of “natural necessity” reappears at III.3 [48] 1.9-10.

34The question is whether such passages license the notion that matter is, for Plotinus, genuinely hypothetically necessary. I think not. Rather, Plotinus seems to be saying that particular examples of evil in our world are hypothetically necessary : hence in the first passage, from II.3 [52], he talks of “evils (τὰ κακά) in the universe” which help make the universe complete. Likewise in both passages from III.2 [47], he is clearly describing the situation that obtains once reason or intellect has already provided the physical world with order and determination. Thus, insofar as Plotinus wishes to identify matter itself with the anangkê of the Timaeus,34 he must have in mind not persuaded, teleologically-ordered necessity, but rather unpersuaded necessity.35 For the production of matter occurs not as the result of any goal-directed enterprise, but “necessarily” in the sense that it is the natural (κατὰ φύσιν) result of the higher principles (III.2 [47] 3.3-5).36 That would leave persuaded necessity to be associated with particular evils in the physical world. These evils are unavoidable concomitants of the order imposed on the physical world by higher causes : that is, they are present in order that the good might be realized as fully as possible. But we need to consider complications on both sides of this neat interpretation, which would associate particular evils with persuaded necessity and evil or matter itself with unpersuaded necessity.

  • 37 Cf. Johansen’s discussion of the difference between Timaean anangkê and mere necessary causes, as d (...)

35First, I consider particular evils : would Plotinus really say that all physical evils are cases of hypothetical necessity ? It may be true that if we are to have a physical world, we will also have evils. But surely hypothetical necessity means more than this. It means that the necessary thing actually contributes to some end. For instance, fuel is hypothetically necessary for operating a car, not just because it is inevitably present when the car runs, but because it plays a role in the running of the car. By contrast, the pollution emitted by the car is not teleological, it is just an unfortunate and unavoidable by-product of the car’s operation.37 Such “by-product” evils may inevitably arise from the pursuit of some telos. But it is the more robust sense of hypothetical necessity, where the necessary thing actually contributes something towards the telos, that seems to be present in our concrete examples of anangkê in Plato. We can name specific purposes for the necessary desires of the Republic, and for instances of necessity in the Timaeus, like the thickness (and thinness) of the skull. It is as yet unclear whether the same is true for every case of evil in Plotinus’ universe.

36He certainly thinks that at least some regrettable phenomena in the physical world are ordered towards a good. For instance, again in the treatise on providence, he writes :

“It would be ridiculous if someone complained of [lower animals] biting men… it is necessary (ἀνάγκη) that these, too, should exist, and some of the benefits that come from them are obvious… so that none of them are in vain (µάτην).” (Enn. III.2 [47] 9.33-7 ; Armstrong trans., adapted)

  • 38 Earlier in the same treatise Plotinus explains that “the occasions where animals set upon and kill (...)

37Another example would be his mention elsewhere of the fact that animals need claws and fangs “as a result of deficiency” (διὰ τὸ ἐνδεές) (VI.7 [38] 9.40). Obviously claws do have a purpose, even if it would in some sense have been better if the animal hadn’t needed them – given that it needs them only to defend itself, to kill other animals, and so on.38 But again, it is not enough that this sort of account be available in some cases of apparent evils or regrettable features of the world. Rather, if particular evils are to be explained as cases of persuaded necessity, such an account should be available for all regrettable features. Plotinus seems aware of this, for shortly thereafter in the same treatise he writes :

“But are all individual things (ἕκαστα) as they are by natural necessities and causal sequences, and excellently disposed in every way that can be (δυνατόν) ? No ; rather reason (ὁ λόγος) makes all these things as their sovereign (ἄρχων), and wishes them to be as they are, and makes the things which are called bad (τὰ λεγόµενα κακά) according to reason, because it does not wish that all should be good, just like a craftsman who does not make everything eyes in his picture. In the same way reason did not make everything gods… not out of a grudging meanness but through a reason which embraces the variation of the intelligible (ἀλλὰ λόγῳ ποικιλίαν νοερὰν ἔχοντι).” (Enn. III.2.11.1-9 ; Armstrong trans., adapted)

38This rather revealing passage shows how apparently pointless evils can be accommodated within a universal teleology. Although a certain evil phenomenon may not serve any obvious immediate purpose, it contributes to an overall proportion and (relative) perfection. It’s worth stressing that Plotinus is not saying simply that the general presence of evils in the world helps perfect it. Rather he is saying that each of the individuals (ἕκαστα) contributes to the goals of the higher, intelligible causes. I conclude from this that Plotinus did hold that all individual evils display persuaded necessity, that is, necessity that is both (1b) goal-directed and (3) regrettable. These are also things that (2b) could have been otherwise, in the sense specified above : it is only in the light of divine providence that we can explain why these things are as they are. Hence Plotinus’ remark at the beginning of the passage that the individuals are not as they are “by natural necessities” (φυσικαῖς ἀνάγκαις). If there is any inevitability involved here, it is not due to the natures of the physical things themselves, but rather to their intelligible causes. And as I have argued, the fact that divine providence may itself be “inevitable” does not prevent (1b) from being satisfied.

39Now for the second part of the interpretation proposed above : is matter to be associated with the unpersuaded necessity of the Timaeus ? This is certainly tempting, especially if it is right that particular evils are to be associated with persuaded necessity. But our discussion of Plato shows that there is an equally tempting rival interpretation, which is that Plotinus understands matter more along the lines of the necessary evil of the Theaetetus. Of course, the difference between the two options is a subtle one : on both, matter/evil would be regrettable and could not be otherwise. The question comes down to whether matter is an ineliminable contrary of the good (Theaetetus), or is merely in itself unpersuaded by the good, and yet susceptible to such persuasion (Timaeus). As we’ve already seen Plotinus quotes both texts prominently in his discussion of evil, and it is not easy to decide which source is more dominant in his thinking. Indeed he might be judged to vacillate on the question. The optimism of the Timaeus comes to the fore when he stresses the impotence and indeterminacy of matter. For instance at Enneads I.8 [51] 13 Plotinus entertains the idea that matter is an “impediment” (ἐµπόδιον) to good (recall that “impeding” is associated by Aristotle with anangkê in his second sense, “compulsion”). Plotinus denies this, arguing that in the case of an evil which is managing to present an impediment, we have something determinate, such as a specific vice or an imperfection in the eye which prevents it from seeing properly. The implication seems to be that the notion of an impediment is already too “positive” a characterization for pure evil itself (cf. the immediately previous distinction between evil and vice in I.8 [51] 12). If even the capacity to inhibit good comes from determination and form, it seems clear that there are no features of matter or the material world that “escape” or stand outside the “persuasive” influence of the intelligible.

40In keeping with this, Plotinus sometimes explicitly claims that matter does entirely obey the higher causes. For instance :

  • 39 This terminology reappears at ch.11 of the same treatise.

“If the maker is prior to the matter, matter will be wholly (πάντη) such as the maker wishes, and easily led39 to all things (εὐάγωγος εἰς ἅπαντα)… when form comes to [matter] it brings everything along with it.” (Enn. II.4 [12] 8.19-24)

41Elsewhere he writes that in fashioning living things, “logos took hold of all matter” (πᾶσαν ὕλην) and brought into matter “all the beings that exist higher up, in heaven” (III.2 [47] 4.18-19). To this we can add the familiar point that Plotinus often seeks to explain evils by saying that they are merely goods that are worse than some other goods :

“A thing which is only slightly deficient in good is not evil, for it can even be perfect on the level of its own nature (δύναται γὰρ καὶ τέλεον εἶναι ὡς πρὸς φύσιν τὴν αὐτοῦ).” (Enn. I.8 [51] 5.6-8)

42Of course such lesser goods are the result of persuasion too ; and as we’ve seen, providence has its reasons for not persuading matter to take on a greater level of perfection in these cases. As we just saw, if the full “variation” of the intelligible is to be reflected in the physical world, not everything can be gods.

  • 40 I.6 [1] 2.16-18 ; I.8 [51] 5.23-4 ; I.8 [51] 9.11-14 ; V.7 [18] 2.15-16.
  • 41 Cf. Porphyry, Sententiae §20.

43Notoriously, though, Plotinus sometimes strikes a less optimistic note, as when he refers to form or intellect failing to “dominate” (κράτειν) matter. This is especially frequent when he is explaining ugliness.40 Clearly the reference in the Theaetetus to evil as the “contrary” of the good plays a role in his thinking here : if matter is the contrary of good then it will somehow resist, despite its indeterminacy. The theme of persuasion from the Timaeus may also be relevant. If we return to the idea that the as-yet-unpersuaded anangkê in the Timaeus is like an audience waiting to be won over by a rhetorician, we may note that the analogy does nothing to suggest the audience initially has false beliefs. But Plotinus says precisely this about matter. He associates it several times with falsehood or lies, as opposed to that which is merely unpersuaded by the truth. For instance he tells us that the soul becomes evil when it reaches opinions along with the body (συνδοξάζουσα, I.2 [19] 3.13), and that “every announcement [matter] makes is a lie” (III.6 [26] 7.21, Armstrong trans.).41 In I.8 [51] itself we find this :

  • 42 Adding µή with Amstrong ; cf. µὴ οὐσία at line 32.

“The evil nature, too, has a false being, primary and absolute falsehood (πρώτως καὶ ὄντως ψεῦδος), but the being of the divine is true being ; so that as falsehood is contrary to truth, so is the non-substantiality42 of the evil nature contrary to the substantial reality of the divine.” (Enn. I.8 [51] 6.44-7)

  • 43 Of course with this whole theme, Plotinus thinking of Republic V and its relation of non-being to f (...)

44This suggests that matter is evil not merely because it is unpersuaded, but because it is actually antithetical to the truth that is to be realized by wise persuasion. Plotinus may, then, have wavered as to whether the intelligible is capable of a complete persuasion of matter. Keeping with the analogy to the case of belief, the question here is whether the falsehood of matter is entirely replaced by true opinions in the physical realm, or whether the natural world displays a mixture of falsehood and truth.43 Plotinus’ ambivalence on the question is well captured in the following passage :

“Therefore, there was need of a concord (ἁρµονία) in which intellect and necessity came together, in which necessity drags it down to what is worse and carries it away to unreason, because it is not a rational principle itself, but all the same, ‘intellect rules over necessity’ (ἄρχοντος δὲ νοῦ ὅµως ἀνάγκης).” (Enn. III.2 [48] 2.33-6 ; Armstrong trans., adapted)

Conclusion : Proclus vs. Plotinus

  • 44 See for instance J. Opsomer, “Proclus vs. Plotinus on Matter (De mal. subs. 30-7)”, Phronesis 46, 2 (...)
  • 45 At in Tim. II.64 he connects the reversion of matter upon nous to Plato’s talk of “persuasion” in t (...)

45I see Plotinus, then, as torn between two understandings of the “necessity” of evil, both of which he finds in Plato. On the one hand, he wants to say that particular evils are necessary in the optimistic and hypothetical sense found in the Timaeus. It is no accident that this dialogue is quoted in the optimistic last remark of the passage just cited. This strand of Plotinus’ thought would identify matter with unpersuaded necessity, which lacks goodness, yet is the occasion for the manifestation of good things in the physical world. On the other hand he is powerfully influenced too by the Theaetetus, where evil is made a necessary contrary to the good. But my reading of Plotinus as pulled in two directions by his Platonic sources was not shared by Proclus, who criticized him for denying that matter “fully reverts” upon the intelligible principles. This is too large a topic to address here, and in any case Proclus’ dispute with Plotinus on the nature of matter and evil has been the subject of some excellent work in recent years.44 For present purposes, it is nonetheless worth noting that the disagreement between Proclus and Plotinus can be framed precisely as one over the sense in which matter is “necessary”. It is not merely that Proclus is more forthright in saying that matter is entirely “persuaded” by nous.45 He also adopts the position I considered, and rejected, as a reading of Plotinus, saying quite explicitly that matter is hypothetically necessary. He writes :

  • 46 For teleology and matter see also e.g. On the Existence of Evils §36.17-37.1 and On the Eternity of (...)

“Matter comes forth from the gods as necessary for the cosmos, and is neither productive of evil (κακοποιός), contributing as it does to the generation of the universe, nor good, being the last of the wholes. Rather, it has its place among those things that are necessary (ἔχει τὴν τάξιν ἐν τοῖς ἀναγκαίοις). For whatever is for the sake of something is like this [i.e. necessary] (πᾶν γὰρ τὸ ἕνεκα τοῦ τοιοῦτον).”46 (In Remp. I 37.30-38.3)

46Because Proclus is so taken with the hypothetical aspect of Timaean anangkê, he is much less emphatic than Plotinus about the fact that anangkê constitutes constraint or grounds for regret. He denies that evil could ever resist being fully “mastered” by form. Rather, everything that exists has form and determination, and evil is just a by-product of the differing levels of such determination. His reading of the very passage in the Theaetetus which underlies the pessimistic strain in Plotinus’ thought is thus strikingly optimistic :

  • 47 Proclus, Tria Opuscula (De providentia, libertate, malo), ed. H. Boese, “Quellen und Studien zur Ge (...)

“Therefore, Plato in the Timaeus is right in saying that in accordance with the will of the demiurge, ‘all things are good and nothing is bad’. In his discussions with the geometer [sc. Theodorus in the Theaetetus], however, he contends that ‘evil things cannot possibly cease to exist’ and that ‘by necessity’ they have come to exist among beings. For all things are made good by the will of the father and, with respect to his productive activity, none of the things that are or come to be are evil. However, when he distinguishes degrees in nature he does not escape the consequence that there is evil for particular things, evil which destroys the good [in them].”47 (On the Existence of Evils §10.1-10 ; Opsomer/Steel trans.)

47It is telling that the Theaetetus citation is framed by a preceding quote from the Timaeus, emphasizing the providential ordering of the cosmos.

  • 48 An added complication here, which I have not addressed, is the role of necessity in the Myth of Er. (...)

48One must not exaggerate the contrast between our two Neoplatonists. For of course Plotinus too argues strongly for the overall goodness of the material world (Enneads II.9 [33]) and the universality of providence (III.2-3 [47-48]). But Proclus clearly emphasizes the optimistic aspects of Plato’s anangkê at the expense of its bleaker connotations.48 His version of necessity is thus arguably less fraught than that of Plotinus. Yet I believe Plotinus comes closer to capturing the tensions and complexity in Plato’s original notion, or notions, of anangkê.

I am very grateful to James Wilberding for his thorough and penetrating comments on an earlier draft of this paper. I also received helpful comments from audiences in Edinburgh and Leuven.

Haut de page

Notes

1 This is taking at face value On Providence §22 ; tellingly Proclus emphasizes the “mortal” nature of the destroyed beams and walls.

2 For a subtle discussion of some of the same themes in a related context, see A. Silverman, “Ascent and Descent : the Philosopher’s Regret”, Social Philosophy and Policy 24, 2007, p. 40-69. Silverman seeks to explain the philosopher’s regret in the Republic context of “descending” to rule the city, but he alludes to the Timaeus account to illuminate this. His paper thus forms a kind of complement to my own, since it uses the Timaeus to help understand the Republic rather than vice-versa. I believe our accounts are largely compatible ; on this see further n.5 below. I am grateful to Prof Silverman for bringing the paper to my attention.

3 G. Van Riel, “Proclus on Matter and Physical Necessity”, in R. Chiaradonna and F. Trabattoni (eds), Physics and Philosophy of Nature in Greek Neoplatonism, Brill, Leiden, 2009, p. 231-58.

4 One reason for the caveat is that the skull, and everything else fashioned by demiurgic activity, will be the best possible imitation of the Forms. Given the demiurge’s goodness and generosity, it “could not have been otherwise” that the skull was of optimal thickness. What I mean by saying that it could have been otherwise is simply that the demiurge and lesser gods do indeed face choices in making physical things : what is necessary is that they always make the right choice. This is not to say that the demiurge and other gods have boundless choice. Rather, they are presumably constrained in a different way by the paradigms : apparently there must be humans and humans must have skulls, and what they are choosing is merely how thick to make the skull. I will return below to the sense in which something like the thickness of the skull “could have been otherwise”.

5 Silverman, “Ascent and Descent”, p. 61-2 and p. 68, differs from me in thinking that regret implies the presence of alternative possibilities, which the philosopher has but the Demiurge does not. But the disagreement may be merely apparent. For Silverman regret implies the presence of superior alternatives. I would agree that these are not present in the Timaeus case, so that in this sense the Demiurge feels no “regret”. In my usage of the term, by contrast, something is regrettable simply if it is sub-optimal, something Silverman does admit (e.g. “[the Demiurge] understands that his creation will be imperfect or, put differently, that his creation will not be fully amenable to the persuasive aspects of reason”, p. 61). And as we have seen there are indeed alternative possibilities for the Demiurge, insofar as he could have pursued some of them at the expense of others, and more generally insofar as he could have fashioned something worse. If this is precluded for him, it is precluded only by his own wisdom, a point I discuss further below.

6 Though the skull is a particularly striking example of this, it is certainly not the only one we are given : for instance at 42a we’re told that souls must “of necessity” be put into bodies and that the bodies’ first capacity will “of necessity” be sensation. (It’s interesting to note that this passage in fact comes before the introduction of necessity at 47e, though Timaeus does say there that his account has mostly (πλήν) dispensed with mentioning necessity.) Later on Timaeus invokes necessity to explain numerous facts about man’s constitution and not just the thickness of the skull. For instance, the fact that muscular parts of the body cannot think (75a), the structure of the digestive system (75d-e), and the corruptibility of man (77a).

7 Here I would follow T. Johansen, Plato’s Natural Philosophy. A study of the Timaeus-Critias, Cambridge University Press, Cambridge, 2004, p. 95, in recognizing “two versions of necessity” in the Timaeus, “the necessity that reason uses as a contributory cause for its end, [and] the necessity that has not yet been persuaded by reason”. In general his account of anangkê is very congenial to my own, although he does not emphasize the regrettable aspect of Timaean anangkê.

8 I take this phrase to be a further explication of the first sense of anangkê ; note it begins kai rather than eti, which marks off each subsequent sense. Thus my numbering of the senses differs from that of Ross.

9 I have here omitted a fourth sense, which has to do with necessity in demonstration.

10 For a comparison of anangkê in the Timaeus to Aristotelian hypothetical necessity, see Johansen, Plato’s Natural Philosophy, p. 115.

11 The point would of course also apply to the further sense of the necessary in demonstration : what follows from contingently true premises will be a contingently true conclusion, even though it follows necessarily.

12 Similarly, to take Aristotle’s sense (ii), if I am forced to act, I might well say to myself that I would have acted otherwise in the absence of the compulsion. This would be a coherent thought even if the compulsion was inevitable.

13 Evenus might be a better guide in this respect, of course. The use of anangkê in tragedy also shows that it has connotations of being “against one’s will” or “regrettable”. Think, for instance, of Aeschylus’ Agamemnon putting on the “harness of necessity”. My thanks to M.M. McCabe for this point.

14 And, interestingly, by means of anangkê itself. The simplest thing to say about anangkê in this passage is that it is being used as a synonym of bia, a use that is mentioned by Aristotle, as we will see. But a more complicated story is possible : from the point of view of the oligarchic man’s epithumetic soul, reason is a regrettable source of constraint. However the epithumetic soul is still able to pursue its own goals, within the limits placed on it by reason. If this is right, reason is a source of anangkê for desire in just the same way that desire is a source of anangkê for reason.

15 An interesting parallel in Republic V labels as a matter of necessity the fact that if they dwell together, women and men guardians will wind up engaging in sexual intercourse (458d) ; Glaucon identifies this as “erotic, not geometric” necessity. Persuasion is used in the political context of the arrangement of the city, to turn this erotic necessity towards an acceptable outcome by carefully regulating this inevitable sexual activity. (By contrast “geometric necessity” would be necessity that lacks these connotations of desire, constraint, and so on.)

16 See G.R. Morrow, “Necessity and Persuasion in Plato’s Timaeus”, Philosophical Review 59, 1950, p. 147-63 ; S.K. Strange, “The Double Explanation in the Timaeus”, Ancient Philosophy 5, 1985, p. 25-39.

17 Strange, “Double Explanation...”, p. 28-9.

18 Arguably the presence of spatial extension is already part of unpersuaded necessity (the pre-cosmic receptacle is, after all, already spatial), like the “shaking” I go on to discuss. But even then it would not be part of unpersuaded necessity that the elements are extended. For without demiurgic activity, there would be no elements, only “traces” of elements. The point may be generalized : in the ordered cosmos, there is nothing spatially extended that is not produced by the gods. (My thanks to Andrew Mason for pressing upon me the claims of the “mechanical” reading of necessity.)

19 Again the Republic might leap to mind, at least for a reader (like a Neoplatonist) who assumes that the doxa vs. epistêmê distinction of Republic V is supposed to correspond to two metaphysical realms, one of particulars and one of Forms.

20 This parallel is also drawn by Van Riel in “Proclus on Matter and Physical Necessity”.

21 Again I agree here with Johansen, who criticizes Grote’s characterization of anangkê as “force… with the negative attribute of not being regular, or intelligible” (quoted at Johansen, Plato’s Natural Philosophy, p. 93). Johansen makes the nice point (p. 99) that necessity can in fact be contrasted to the bia mentioned in the context of the creation of the world soul at Timaeus 35a.

22 On this treatise see A. Schniewind, L’Éthique du sage chez Plotin : le paradigme du spoudaios, “Histoire des doctrines de l’Antiquité classique 31”, Vrin, Paris, 2003, and K. McGroarty, Plotinus on Eudaimonia. A Commentary on Ennead I.4, Oxford University Press, Oxford, 2006.

23 For a similar use of anangkê see II.9 [33] 9.11.

24 A related use of anangkê can be found in Plotinus’ treatment of fate (heimarmenê), which he describes as a case of anangkê that arises for humans insofar as they are tied to the physical world. In II.3 [52] 9, Plotinus mentions fate as part of his treatment of the Myth of Er, and says that whenever we submit ourselves to physical necessitation (he mentions the “necessary affections” of Timaeus 69 at II [52] 3.9.8), we live “within fate” (II.3 [52] 9.28). A courageous man, by contrast, imposes himself on the situation presented to him by fate. The situation is not “up to him”, but his response to the situation is. This contrast is made explicitly at IV.3 [52] 15.10-15, where Plotinus says that whereas some people subject themselves completely to fate, others accept “necessary things” (ἀναγκαῖα) but still manage to perform actions that are their own.

25 The cases may not be exactly analogous. For one thing the sage is really the agent of virtuous deeds, whereas he is not really his body : the sage “is self-sufficient for happiness and the acquisition of good ; for there is no good that he has not got. What he seeks he seeks as a necessity (ὡς ἀναγκαῖον), not for himself but for something that belongs to him ; that is, he seeks it for the body which is joined to him.” (I.4 [46] 4.23-7 [Armstrong trans., adapted])

26 As in I.4, Plotinus stresses that although these teloi are accepted by the sage, they are not to be identified with eudaimonia itself.

27 For this treatise see D. O’Meara, Plotin. Traité 51, “Les écrits de Plotin”, Les Éditions du Cerf, Paris, 1999. See also his “Das Böse bei Plotin (Enn I,8)”, in T. Kobusch and B. Mojsisch (eds), Platon in der abendländischen Geistesgeschichte, Wissenschaftliche Buchgesellschaft, Darmstadt, 1997, p. 33-47.

28 The literature on evil in Plotinus is of course vast and contentious. A fundamental study is D. O’Brien, “Plotinus on Evil. A Study of Matter and the Soul in Plotinus’ Conception of Human Evil”, in Le Néoplatonisme, “Colloques internationaux du CNRS”, Éditions du CNRS, Paris, 1971, p. 113-46. For a book-length treatment of O’Brien’s position, see his Théodicée plotinienne, théodicée gnostique, “Philosophia antiqua 57”, Brill, Leiden, 1993, and the discussion of it in G. Leroux, “Métaphysique et théodicée chez Plotin : remarques sur les travaux de Denis O’Brien”, Dialogue 35, 1996, p. 293-306. See also e.g. J.-M. Narbonne, La Métaphysique de Plotin, Vrin, Paris, 1994 ; L. Lavaud, D’une métaphysique à l’autre. Figures de l’altérité dans la philosophie de Plotin, Vrin, Paris, 2008 ; and J.M. Rist, “Plotinus on Matter and Evil”, Phronesis 6, 1961, p. 154-66. For Plato as a source of the Plotinian position see R.W. Sharples, “Plato, Plotinus, and Evil”, Bulletin of the Institute of Classical Studies 39, 1994, p. 171-81.

29 Note that Plotinus replaces οὔτ᾿… δυνατόν with simply µή, one of several variations from the original in this quotation. These variations are noted, but more or less dismissed as a case of Plotinian carelessness, by E. Schröder, Plotins Abhandlung Pothen ta kaka (Enn. 1,8), Ernst Borna, Leipzig, 1916, p. 153. But see also O’Meara, Plotin. Traité 51, p. 124. One notable change is from hupenantion (“sub-contrary”) to the more usual enantion (“contrary”) at the end of the passage. This is significant, given the emphasis Proclus will later place on the unusual terminology. The present passage is not the only one in which Plotinus alludes to Theaetetus 176 ; see for instance I.2 [19] 1, III.2 [47] 5, III.2 [47] 15.

30 For Plotinus’ argument to this effect see D.J. O’Meara, “The Metaphysics of Evil in Plotinus : Problems and Solutions”, in J. Dillon and M. Dixsaut (eds), Agonistes : Essays in Honour of Denis O’Brien, Ashgate, Aldershot, 2005, p. 179-85. O’Meara’s analysis is criticized in J. Opsomer, “Some Problems with Plotinus’ Theory of Matter/Evil. An Ancient Debate Continued”, Quaestio 7, 2007, p. 165-89.

31 By contrast to the Theaetetus quotation this is fairly accurate, with the notable exception of substituting “φύσις” (nature) for “γένεσις” (offspring or generation).

32 L. Gerson, Plotinus, Routlege, London, 1994, p. 192-7, also invokes hypothetical necessity in this context. He has been criticized for this by O’Meara (“Das Böse bei Plotin”, p. 11 n.26), who cites I.8 [51] 7.2-6 to show that the necessity Plotinus has in mind is not hypothetical.

33 This is a rather difficult passage, situated as it is in the midst of a dialectical consideration of strife in the physical world. But I take it that Plotinus is endorsing the sentiments that all is as good as it can be, that matter is caused by logos for the sake of the world being as it is, and so on. The point about the argument being unhelpful is simply that this does not yet explain how a single logos can produce internal conflict.

34 Though it is common in the secondary literature to ascribe this claim to Plotinus, as far as I can tell it is not one he ever makes fully explicit. Still, it does seem to follow from such texts as I.8 [51] 7.1-5 and III.2 [47] 2.33-40. Enn. III.8 [30] makes explicit remarks about evil and matter as relating to Plato’s Timaeus. On this see B. Collette-Ducic, Plotin et l’ordonnancement de l’être, “Histoire des doctrines de l’Antiquité classique”, Vrin, Paris, 2007, ch.3.

35 James Wilberding has suggested to me a compelling argument against identifying matter with persuaded necessity. If persuaded necessity is both hypothetically directed towards a goal and regrettable, then it should be possible to explain this by referring to different features of what is necessary. For instance taking medicine is regrettable insofar as it is bitter, but goal-promoting insofar as it cures an illness. Since matter is simple and has no multiplicity of features (indeed, no features at all), it cannot be both regrettable in one way and goal-promoting in another.

36 The idea of “natural necessity” reappears at III.3 [48] 1.9-10.

37 Cf. Johansen’s discussion of the difference between Timaean anangkê and mere necessary causes, as described in the Phaedo (Johansen, Plato’s Natural Philosophy, p. 104-5). His example of eye color (at p. 100-1), taken from the Aristotelian context, would be similar to my pollution example, though it leaves out the idea that the necessary is regrettable.

38 Earlier in the same treatise Plotinus explains that “the occasions where animals set upon and kill one another are necessary, for they were not generated as eternal” (οὐδὲ γὰρ ἀίδια ἐγίνετο) (III.2 [47] 4.16-17).

39 This terminology reappears at ch.11 of the same treatise.

40 I.6 [1] 2.16-18 ; I.8 [51] 5.23-4 ; I.8 [51] 9.11-14 ; V.7 [18] 2.15-16.

41 Cf. Porphyry, Sententiae §20.

42 Adding µή with Amstrong ; cf. µὴ οὐσία at line 32.

43 Of course with this whole theme, Plotinus thinking of Republic V and its relation of non-being to falsehood and being to truth. For him, non-being is matter, being is the intelligible, and this leaves particulars to be the realm of opinion.

44 See for instance J. Opsomer, “Proclus vs. Plotinus on Matter (De mal. subs. 30-7)”, Phronesis 46, 2001, p. 1-35.

45 At in Tim. II.64 he connects the reversion of matter upon nous to Plato’s talk of “persuasion” in the Timaeus.

46 For teleology and matter see also e.g. On the Existence of Evils §36.17-37.1 and On the Eternity of the World §11. For the Republic commentary, see Proclus, In Platonis Rem publicam commentarii, ed. W. Kroll, 2 vols Teubner, Leipzig, 1899. For a French translation see Proclus, Commentaire sur la République, trad. par A.J. Festugière, 3 vols, Vrin, Paris, 1970.

47 Proclus, Tria Opuscula (De providentia, libertate, malo), ed. H. Boese, “Quellen und Studien zur Geschichte der Philosophie I”, W. de Gruyter, Berlin, 1960. For the three opuscula, the Latin text and Greek excerpts by the Byzantine author Isaac Sebastocrator are also printed, with French translation, in Proclus, Trois études sur la providence, ed. and trans. D. Isaac, 3 vols, Les Belles Lettres, Paris, 1977-82. For English translations see Proclus, On the Existence of Evils, trans. J. Opsomer and C. Steel, “Ancient Commentators on Aristotle”, Duckworth, London, 2003, and Proclus, On Providence, trans. C. Steel, “Ancient Commentators on Aristotle”, Duckworth, London, 2007.

48 An added complication here, which I have not addressed, is the role of necessity in the Myth of Er. Proclus identifies anangkê in this setting as an intellective god who “rules over all things in the cosmos” (τὴν [ἀνάγκην] πάντων ἄρχουσαν τῶν ἐγκοσµίων) (in Remp. II 206.2-3). He argues explicitly that this anangkê is not to be identified with the anangkê mentioned in the Timaeus.

Haut de page

Pour citer cet article

Référence papier

Peter Adamson, « Making a Virtue of Necessity : Anangkê in Plato and Plotinus », Études platoniciennes, 8 | 2011, 9-30.

Référence électronique

Peter Adamson, « Making a Virtue of Necessity : Anangkê in Plato and Plotinus », Études platoniciennes [En ligne], 8 | 2011, mis en ligne le 16 décembre 2014, consulté le 29 avril 2017. URL : http://etudesplatoniciennes.revues.org/442 ; DOI : 10.4000/etudesplatoniciennes.442

Haut de page

Droits d’auteur

© Société d’Études platoniciennes

Haut de page
  • Revues.org