Comments on Julia Annas, Platonic Ethics, Old and New
Tim O'Keefe
Minnesota Ancient Philosophy Conference, April 19, 2003
The main task of Julia Annas' Platonic Ethics, Old and New is to question some old (or maybe new!) orthodoxies and to spur debate. She wants to rouse us from our developmental slumbers by drawing our attention to the very different way Plato was read by the group of thinkers we nowadays label the 'Middle Platonists.' Her book succeeds marvelously in that task-it has already opened debate on the wide range of issues in Plato she addresses.

Today's session is a fitting example of the sort of reassessment of Plato spurred by Annas' book. In my time here, I want to concentrate on one aspect of the book'chapter 5, 'What Use is the Form of the Good? Ethics and Metaphysics in Plato.' I'll start out by summarizing Annas' argument against the modern view that Plato's ethics, at least in the Republic, depend on his metaphysics-particularly on the form of the Good. Then I'll set out her reasons for endorsing the Middle Platonist view that ethics and metaphysics are distinct areas of Plato's philosophy, without one being dependent on the other. Finally, I'll raise some questions of my own about her arguments and her account.

Initially, it may seem obvious that the Republic's ethics depends on its metaphysics. After all, Plato states that the guardians, who have expert knowledge of what is good and just, must have knowledge of the form of the Good. (506a) It's through imitating the forms' order and beauty that the philosopher is himself able to become as divine and ordered as a human can (500b-c), and by looking to the forms that he knows how to order to city and its citizens (501b; cf. 520c). It's by their relations to the form of the Good that other things become useful and beneficial, and if we don't know the form of the good, even the fullest knowledge of other things is of no benefit to us. (505a) Finally, the authority of the guardians to direct the lives of the other citizens as they see fit, whatever the citizens may wish, is based upon the guardians' grasp of the form of the Good.

However, says Annas, none of this shows that the particular ethical content of the Republic's ethical theory depends crucially on its metaphysics. After all, the 'moral authoritarianism' of the Republic is equally present in the so-called 'Socratic' dialogues, where Socrates asserts that people with expert knowledge of the virtues are thereby justified in ruling over the ignorant. Likewise, in those dialogues, expert knowledge of the virtues is necessary and sufficient for happiness. In both cases, it's expertise that's central, not forms. There may be a change from the 'Socratic dialogues' to the Republic concerning what's required for expertise, with the Republic describing an elaborate education system needed to get in contact with transcendent Forms and thus gain expertise. However, the Forms answer 'specific epistemological and metaphysical problems' as Annas puts it, 'the development and separation' of Forms'applies to the cases of the double and the half just as much as to the good and the bad.' (98) and don't greatly effect the specifically ethical content of the Republic, which is similar to what's put forward in so-called 'earlier' dialogues innocent of forms.

Despite this, somebody might say that the metaphysics grounds the ethics in the way that theory grounds a skill or expertise; i.e., it's through knowledge of the forms that one gains the requisite ethical expertise. But there are two problems with this, says Annas:

First, the Republic doesn't have an actual metaphysical theory, only a sketch of one, whereas the ethical theory is presented in a perfectly straightforward, confident way. So it seems that the metaphysics isn't grounding the ethics in the Republic.

Secondly, if the metaphysics is supposed to ground ethics, then the following problem arises: 'virtue turns out to be self-undermining: understanding its basis turns you against practicing it.' (p. 105) That's because Plato's emphasis on the joys of abstract thinking makes any return to mere practice seem worthless by comparison.

Annas also considers and rejects several other possible ways in which it might be thought that Plato's ethics depends on his metaphysics, but I'll return to these later.

Because of these problems with our modern way of viewing the relationship of Plato's ethics to his metaphysics, Annas suggests that we can profit by considering the Middle Platonists. They thought that 'Plato's philosophy should be approached in terms if the three parts of philosophy: physics, ethics, and logic.' (108-9) These later Platonists assimilate Plato to the Stoics' systematic division of philosophy into parts, and the Stoics use various metaphors to illustrate the relationships these parts have to one another, e.g., philosophy is an egg, with logic its shell, ethics the white, and physics the yolk.

Annas says this is a non-hierarchical view of the parts of philosophy, with no part having primacy over the others: the parts 'hold together in a distinctive and mutually interdependent way.' (111) She reports Apuleius' view that the 'needed parts would not only not conflict, but even help one another with mutual aid, forming an organic whole'' (111)

Annas claims that this non-hierarchical view 'excludes some views of the relation of metaphysics to ethics and allows and encourages others. On this view, ethics cannot depend on metaphysics for its content, or it would not be a distinct part. To attempt to derive ethical conclusions from metaphysical premises would be like trying to derive conclusions in logic from metaphysical premises; it would be to confuse two distinct philosophical tasks. Ethics has to be developed on its own.' (111-2)

But she also says that ethics 'has to be seen in the context of' metaphysics, and vice-versa.

We will, for example, understand human rationality better if we see it not as an isolated ethical phenomenon comprehensible only in our own lives, but as an example of something visible on a bigger scale, in the regularities of the world and of the heavenly bodies. By having our understanding of rationality expanded in this way, we can become inspired to become more rational ourselves. We have not been given cosmic premises which mysteriously direct us to act in one way rather than another; rather, we have been given a picture of ourselves as rational beings in a universe where rationality is dominant, and this alters our conception of ourselves by deepening it and putting it in context. (112)
Viewed in this way, 'Ethics is a distinct part developed with a cluster of eudaimonistic concepts which are independent of metaphysical concepts.' (115) Metaphysics might help to 'stabilize' one's ethics, as she puts it, and give one's ethics a 'context,' but ethics can't depend for its content on metaphysics. Annas introduces the metaphysics/ ethics gap, analogous to the is/ought gap or the fact/value gap: 'to think that ethical conclusions can be obtained from metaphysical premises is thus to be in a muddle about what ethics and metaphysics are.' (115) [NB: this is my term, not Annas'! In discussion afterwards, she indicated that she didn't want to be credited with this doctrine.]

To reinforce this metaphysics/ethics gap, Annas points to the similarity of the ethics of Plato and the Stoics, despite their radically different metaphysics. One could move from being a Platonist to a Stoic without thereby fundamentally altering one's ethics. Annas draws a similar lesson from a surprising place: the discussion of the existence and nature of the gods in Book 10 of the Laws, where stringent penalties for atheism and heterodoxy are set forth in order to safeguard the citizen's ethical views. Annas points out that the traditional Olympian views of the gods previously adequately grounded the ethical views of the citizens, and it's only when challenged by the sophists that a new conception of the gods needs to be brought in and defended: divinity as cosmic reason governing the kosmos. Annas says that this shows that 'ethical beliefs clearly cannot be derived from the metaphysical background, since the person has to have the ethical beliefs already before needing the metaphysics.' (113) Furthermore, 'the rational metaphysical beliefs [cannot] in any obvious sense justify or ground the ethical beliefs, since whatever it is that they contribute was previously contributed by the implicit metaphysics of traditional religious beliefs.' (114) Both traditional and rational theologies might 'provide a context which makes sense of ethics and thus renders the person's ethical beliefs more stable and reliable,' but that's quite different from having ethics depend on metaphysics.

Initially, I'm suspicious of Annas' proposed 'metaphysics/ethics gap.' Since ancient ethicists do not generally subscribe to the fact/value gap, our presumption should be that facts about the way the world in general is structured, what sorts of entities exist, and the like'that is, metaphysical premises'should be able to have an impact on the conclusions one reaches about ethical questions such as what the best life is for a human being and how to attain it. I don't see how this involves being in a muddle about what metaphysics and ethics are.

The Middle Platonist division of Plato's philosophy doesn't support Annas' 'metaphysics/ethics gap.' Most of the Middle Platonist writings we have are compendia of Plato's putative 'doctrines'; for such compendia, it would be natural to set out Plato's thought systematically in terms of the different subject-areas of philosophy. Nothing follows one way or the other about the dependence or lack thereof of the different sub-divisions to one another. None of the Middle Platonists, as far as I know, set forth anything like the 'metaphysics/ethics gap;' the best Annas can offer here is an argument from silence.

I draw a different lesson from the metaphors used to describe the parts of philosophy than does Annas. Annas says that metaphors such as that of the egg conjure up a non-hierarchical view of the relationship between the branches of philosophy, and that therefore, ethics can't depend on metaphysics for its content.

But this doesn't follow. If A and B are interdependent, as Annas puts it, of course A can depend on B, as long as B in turn also depends on A. The metaphors suggest the following position: the parts of philosophy are interdependent in a way such that, although distinct, they can't be separated and understood in isolation-part of what makes an egg white an egg white, with its characteristics, is its functional role in the egg as a whole and its relationship to the other parts, although of course the white isn't the same thing as the yolk or the shell.

Annas' further argument, that the Stoics were able to come to many of the same ethical conclusions from a very different metaphysical system, and that the traditional religion of the Athenians was able to ground their ethics as well as the rational religion Plato argues for, is inconclusive. That one can reach the same ethical content without presupposing the same metaphysical system doesn't prove anything. It could still be the case that, for Plato, the ethical conclusions depend on the metaphysical doctrines. We can draw a parallel with arguments. That the same conclusion can be reached from multiple premise sets doesn't show that, within some particular thinkeros philosophical system, that conclusion doesn't depend on some set of presuppositions. Annas seems to be working with an overly strong sense of 'dependence' here, according to which ethics would 'depend' on metaphysics only if the metaphysical presuppositions were necessary conditions on the ethical content. But even given this sense, her position does not follow. Stoics, or ordinary citizens, might believe all sorts of ethical doctrines that may overlap with Plato's ethics. Plato, though, can deny that they have adequate justification for believing these ethical theses.

Certainly, people innocent of metaphysics can have ethical beliefs; for instance, that virtue has a type of value different than ordinary goods. But if we start from the endoxa and try to make sense of them, we'll soon be led beyond ethics, to a consideration of what the world must be like for these ethical beliefs to hold. If this picture is right, we can read Plato's discussion of rational theology in book 10 of the Laws quite differently than Annas does. The Athenian admits that the unreflective religious beliefs and practices of the crowd can 'ground' ethical behavior, in some sense of 'ground.' But I doubt that Plato would assent to Annas' claim that this shows that ethical beliefs cannot be derived from a metaphysical background or that 'the rational metaphysical beliefs [cannot] in any obvious sense justify or ground the ethical beliefs, since whatever it is that they contribute was previously contributed by the implicit metaphysics of traditional religious beliefs.' (114) This is for two reasons. First: the traditional religious beliefs can't be rationally justified, since they are incoherent: for instance, they show the gods both as paradigms of virtue and as squabbling, jealous, and bribable. These unjustifiable beliefs could not themselves justify or ground correct ethical views as a rational theology can. Secondly, even if the traditional theology were justifiable, it wouldn't provide cogent grounds for correct ethical views. Although unreflective conventionalists might swallow the traditional morality that comes packaged alongside traditional religion, traditional religion contains elements that, for a reflective person, undermine traditional morality. That's because the traditional theology sometimes displays the gods as bribable, jealous, etc., and Plato makes it clear (e.g., in the Euthyphro, Republic 2 and 3, and the Laws) that these beliefs about the gods encourage vicious behavior. So traditional unreflective religious cannot ground ethical beliefs in the same way as can Plato's rational theology, and the mere fact that ignorant and unreflective people may have some correct ethical beliefs as a result of their upbringing doesn't show that Plato thinks that ethics can't depend on metaphysics.

If Plato did subscribe to the 'metaphysics/ethics gap,' it would be hard to understand his insistence that harmful ethical conclusions would be obtained from incorrect metaphysical premises. In Laws 889a-890a, the Athenian describes a mechanistic worldview: the workings of the universe are due to 'nature and chance' (principles like the hot, the dry, etc.), as opposed to a teleological view, according to which they're due to art and intelligent planning. Although the Athenian later generously admits that there may be some naturally just, good-hearted atheists (908b), he makes it clear that what's really implied by the mechanistic theory (889a) is that societal moral standards are merely conventional ('goodness according to law'), with a concomitant adherence to a Calliclean 'goodness by nature,' where force and flouting these conventions are justified if needed to gratify one's desires. (I believe that Plato holds (incorrectly) at this is the ethical position actually implied by the mechanistic premises, not merely that a mechanistic world-view tends to have the psychological effect of predisposing certain people to Callicleanism. The latter is true, but it is to be explained because of the former being true.)

Annas might agree with much of this, although I'm not certain. She writes that ethics has to be viewed 'in the context of metaphysics,' and vice-versa. Having correct metaphysical beliefs can help 'make sense of ethics' and thus 'render the beliefs more stable and reliable.' I'm unsure exactly what she means by these phrases, but the most straightforward way that correct metaphysical beliefs can accomplish these things for ethical beliefs is by justifying them-given that p, q and r are true of the world, my ethical belief e 'makes sense' and is 'more stable' because e but follows from p, q and r, perhaps in conjunction with other beliefs, and thus e is now knowledge-or at least e has a higher epistemic status than it did previously. But if this is right, then ethics does seem dependent on metaphysics in a way-if the two were totally autonomous, how could a change in metaphysical beliefs have any impact whatsoever on one's ethics, making it more stable, reliable, or sensible?

However, I gather that the relationship Annas envisions is weaker than this. She writes on the ethical role played by appealing to cosmic reason, and viewing one's own reason as a microcosm of the macrocosmic reason that structures the world. She says that looking to reason working in the universe at large can provide a 'model' for us to emulate; doing so can also impress us and inspire us to become rational ourselves. (107) These ethical functions, although useful, seem rather peripheral-the model may help some people, but it isn't necessary for the assiduous, and looking to Cosmic Reason might help inspire us or buck us up when we feel fainthearted, but this hortatory function isn't necessary; if Cosmic Reason didn't exist, it seems, this wouldn't have any important ethical impact.

I may be stressing the autonomy of ethics and the 'ethics/metaphysics gap' in forms stronger than Annas actually wants to endorse. [NB: Annas said afterwards that this was so.] If so, I'd like to hear what exactly the principle is that she wishes to ascribe to Plato, and whether it would allow for the sort of justification of ethical principles in terms of metaphysics that I've described above.

I've been explaining my doubts about Annas' arguments, and discussing, at a fairly high level of generality, the ways in which I think Plato might think that ethics can be dependent on metaphysics. But let me get down to brass tacks now, and briefly describe three specific cases in which particular ethical doctrines of Plato are dependent on metaphysical doctrines. In each case, I will allege, some of the detailed content of Plato's ethics depends on his moral psychology, and his moral psychology depends on his metaphysics. Hence, at least indirectly, some of Plato's ethics depends on his metaphysics.

  1. Moral education in the Republic. Annas asks: how could the epistemology and metaphysics of the Republic make a difference to the content of what the philosophers hold to be true about value? Annas herself gives one answer to this: because Plato assimilates learning about the virtues to the model of learning mathematics, which is based upon the nature of the forms as transcendent objects, people have to engage in all sorts of abstract and mathematical thinking in order to find the truth. So here's one difference in the content of the ethical theory that results from Plato's metaphysics: mathematics and other such subjects will be good things to study, because studying them helps develop the habits of abstract thinking, divorced from dependence on the senses, necessary for grasping the Forms and hence becoming wise and good. By contrast, Diogenes the Cynic and Epicurus, because of their differing epistemologies and metaphysics, don't think too highly of math and the like. Diogenes denounces such study as utterly irrelevant and perhaps even distracting and harmful.

    After noting that the Republic does change what is required to come to know the truth, however, Annas asks, 'does this alter the content of the moral theory that the virtuous person exemplifies and understands?' Annas might be thinking that the Republic's educational program concerns only how one becomes virtuous, which isn't part of the content of the ethical theory as such. But this distinction is artificial. I'd ask then what exactly it means to alter the content of an ethical theory based on metaphysical and epistemological considerations. If followed consistently, Annas would have to say that Aristotle's discussion of habituation and the like isn't properly a part of the content of his ethical theory. But if questions about how to develop a virtuous character aren't a part of ethical theory for the ancient Greeks, what is?

  2. Attitude toward the beloved in the Symposium. In the Symposium, eros is conceptualized as a desire for the Beautiful Itself, the possession of which will bring blessedness. Ascent toward the form of Beauty can start with erotic desire for a particular beautiful boy. If pursued rightly, however, you recognize the likeness of the beauty of the boy's body with the beauty of other bodies, and then you go on to recognize the beauty of laws, souls, bodies of learning, etc., until you finally reach the vision of Beauty Itself. The metaphysical doctrine that the transcendent form of Beauty is the cause of whatever beauty particular beautiful things instantiate, and the associated psychology of eros as desire for this form, has at least two ethical consequences. The first is similar to what I just said about the Republic: the process whereby you become virtuous and happy depends on the metaphysics. eros is a valuable means for attaining blessedness, because it is an impulse toward the Beautiful Itself. However, it must be directed rightly, so that you don't wallow in sensual gratification; instead, you have to be prepared to abstract from your desire for the particular beautiful boy and recognize beauty in all its instantiations, particularly 'abstract' ones. Secondly, the metaphysics has an impact on the beloved's value and the correct attitude toward him. The beloved is valuable, but not because of his particularity'instead, he is valuable insofar as he instantiates Beauty Itself and other valuable Forms, and thus is valuable for the sake of the ascent. I won't get into the heated issue of whether Plato's view of eros is repugnant, or insightful, or both, but I think this is certainly a difference in ethical content. I could make very similar claims that the metaphysics and associated moral psychology of the Phaedrus, where the lover does not recognize that recollection of the forms is what truly drives him forward, effects its ethics.

  3. Impersonal demands of the Good in the Republic. The final way in which Plato's metaphysics affects the content of his ethics is also brought up by Annas herself. She notes that one of the important themes of the central books of the Republic is that virtuous people 'come to comprehend that virtue is not simply in their own interest, but makes demand on them, to which they respond, whatever their interests.' Besides what is good for some individual, or for some state, there is such a thing as the Good as such, or the Good itself, which the philosopher comes to recognize and to respond to in his training in abstract thinking. (102)

    Annas, however, disputes this conclusion. She says that the Stoics also think that virtue makes this sort of impersonal demand, even though they utterly reject Platonic metaphysics and epistemology. As I noted before, even if they did utterly reject Plato's metaphysics, this on its own is inconclusive, since different thinkers may believe that their metaphysics grounds their ethics in different ways.

    But in any case, I think that the metaphysics of Plato and the Stoics aren't far apart, at least as far as the parts of their metaphysics that allow them to think that virtue makes an impersonal demand, and that the good person could recognize this sort of demand. It's not the form of Good as such that is central here, but a different metaphysical point that Plato has in common with the Stoics'as Annas puts it, 'both believe in a teleologically ordered universe organized by cosmic reason'; value is fundamental to explaining why the world is the way it is.

    This notion of 'cosmic reason' that the two share Annas calls a 'quite different model' of a metaphysical background to ethics' from the Republic's appeal to the Forms, but within Plato's thought I don't believe that the 'two models' are much different at the end of the day. After all, in the Timaeus, cosmic reason, when working to order the world rationally, uses its knowledge of the forms in shaping matter to be as ordered as it can be. And in the Phaedo 97c-98b, Socrates says that the most satisfactory sort of explanation would be learning the final causes of all things, how they are each fashioned by Mind 'in the way that was best.' (97c) However, he's unable to find this kind of cause (99c-d). So, as a second best, (99d, 100b) he turns from final explanations to formal explanations, saying, e.g., that things are good because they participate in the form of the Good. So I do not think that realism about universals as such is key to the ethics'otherwise, David Armstrong's ethical views should be like Plato's, and Annas is correct that Plato's view on universals as such is simply a metaphysical point that applies as much to the double and half, as to the good. Instead, what matters is that there are value universals' mind-independent objective values'that are fundamental in explaining why the universe is structured in the way that it is. As John Leslie, a modern-day Platonist, puts it, 'the universe exists because it ought to.'

    Annas points out this similarity in Platonic and Stoic metaphysics, but then she says that this 'is also manifestly not why they can agree so extensively in ethics. That comes from their convergence in discussions framed in eudaimonistic terms. Neither the agreements nor disagreements in metaphysics affect their agreement that virtue has a value different in kind from that of the conventional goods.' (115) But I'm not so sure. Without the metaphysical assumptions shared by the Platonists and Stoics, how do we explain the way in which virtue makes impersonal demands? The example of the Epicureans is instructive at this point. They vehemently disagree with the Stoics and Platonists that there is any such thing as what is good as such. Polystratus, the third head of the Garden, says 'good' is always a relational term, like 'bigger' or 'healthy'; to think that there can be something good per se, not good for somebody, is to make a category mistake, like thinking that something can be bigger per se. But, as is often the case, metaphysical considerations are lurking behind what is ostensibly a piece of semantic and linguistic analysis. The Epicureans' mechanistic world-view rejects the sort of teleological world-ordering that can ground an impersonal good as such. Annas says: 'No change in your ethical views therefore seems called for by a radical change in metaphysical views.' This may be true if moving from a Platonic to a Stoic system'but that's because, despite their disagreement about universals, they agree on the key metaphysical point. But what about moving to an Epicurean world-view? Plato, it seems, thinks that this sort of change would alter your ethical views, as his discussion in the Laws makes clear.

None of this implies that particular ethical prescriptions can be deduced from quite abstract metaphysical considerations-you don't conclude that because there is a form of Justice Itself, you shouldn't return an axe to your insane friend. Cosmic premises do not mysteriously direct us to act one way rather than another. However, I think that, for Plato, central ethical theses are supported by appealing to metaphysical considerations, in the sorts of ways I've sketched above. Certain ethical theses might initially be suggested by eudaimonistic considerations from within ethical thinking, but in the end ethics is not independent of metaphysics: ethical conclusions can be grounded in, or undercut by, metaphysical considerations. I think that Aristotle can be a useful analog here, although what I'm about to say about him is controversial. Within Aristotle's system, some of his ethics does depend quite a bit on his metaphysics. Because there is a human phusis with its own telos, which can then be used as the standard of flourishing, Aristotle is able to deploy the function argument in order to have an ethical system that's naturalistic and teleological without thereby being subjectivist. Other thinkers can develop ethical systems that share much of the content of Aristotle's ethics without buying into his metaphysical biology, but I think that, for Aristotle, his ethics has an important metaphysical basis.

Nor does what I am saying imply that ethics is subservient to metaphysics, since metaphysical conclusions can be likewise deduced from or supported by ethical considerations. I think Plato would say that one reason to reject the mechanistic world-view described in the Laws is because the Calliclean ethics implied by it is unacceptable. To say that ethics and metaphysics can influence each other in this manner jibes better with Apuleius' view that the parts of philosophy give one another mutual aid than does Annas' assertion that ethics is developed quite on its own, independently of metaphysics.

Although I've drawn rather different lessons from the Middle Platonists than Annas does on the topic of the relationship between ethics and metaphysics in Plato, I agree with her that we have a lot to learn from them, on a wide variety of issues within Plato's writings. We should be grateful to Annas for helping draw attention to these neglected figures. By introducing their voices to us, she has helped enrich our understanding of Plato.

Back to the list of Tim's publications and presentations