The Studia Philonica Annual 15 (2003) 35–53 THE STOICS AND AQUINAS ON VIRTUE AND NATURAL LAW* Phillip Mitsis I wish to begin with a few general observations about the natural law tradition and the striking thesis defended by Paul Vander Waerdt and Brad Inwood about its origins in antiquity. In Vander Waerdt’s more extreme formulation, the early Stoics, whom he believes originated the theory of natural law, dispensed ‘entirely with rule-based reasoning’ and proposed a ‘dispositional model of natural law which prescribes not the external characteristics of morally correct actions but rather the intensionalist features of the agent’s motivation which guarantee that all of his actions are morally infallible.’1 Natural law, in this view, is strictly isomorphic with the perfected rational disposition of the sage, and it prescribes nothing over and above just what such a perfectly virtuous agent should be internally disposed to do on any given occasion. Thus, at the heart of the thesis is the claim that the theory of natural law originates as a kind of radically particularist moral theory that does not presume to offer moral guidance of a general nature and that has as its exclusive focus the inner virtuous dispositions of moral agents engaged in discrete episodes of moral choice. Presumably, one might think that in principle it would be possible to derive more general, law-like prescriptions for moral actions from the reasoning behind them or from the rational and moral dispositions that engender them. But the claim here is that both moral reasoning and moral disposi- tions themselves are not susceptible of law-like analysis and that they are manifested in a series of discrete behaviors that themselves are immune to * I am indebted to Norman Kretzmann for first introducing me to these issues and for many years of subsequent conversation. Paul Vander Waerdt and Brad Inwood have been defending an opposing view to the one set out here about Stoicism for so long and so ably that the present paper may seem to reflect merely a bad case of intellectual stubbornness. In any case, I wish to thank them for their friendship and intellectual indulgence. I am also extremely grateful to Scott MacDonald for comments on an earlier version of this paper, and to audiences at Columbia, Oxford, London, Geneva, Bern, Kyoto, and Notre Dame for many helpful observations. 1 This volume, p. 18. Inwood’s fullest statement is to be found in (1999) pp. 95–127. He argues for a ‘more procedural understanding of moral ‘law’’ wherein ‘‘law’ represents the prescriptive force behind the correct moral choice of the ideal moral reasoner, the sage, whatever the content of that choice might be on a given occasion.’ (pp. 96–97). Thus, his account differs from Vander Waerdt’s in important ways. I will be discussing various aspects of both accounts as I go along. 36 phillip mitsis more contentful general analysis. One is probably more likely to associate the kind of radical particularism claimed here for the Stoics with intuitionist theories of moral epistemology, rather than with their own relentless emphasis on rationality, but the thesis yokes together other strange philo- sophical bedfellows. Another crucial assumption in this thesis about the origins of natural law is its postulation of a strong link between moral infallibility and particularism. It holds that since any non-vacuous moral rule or general recommendation inevitably falls prey to exceptions, the moral deliberations of infallible moral agents cannot be guided by generali- zations at any level. Again, one might be more used to associating claims about the inevitable failure of moral generalizations with theories that dis- play a heightened emphasis on the indeterminacy of moral choice and the inexactness of the moral domain, and hence, presumably, on the conse- quent fallibility of moral agents. But perhaps the greatest paradox of all is the central claim of the thesis itself. Surely, it is prima facie a little paradoxical to claim that the tradition of natural law begins as a theory rooted in the assumption that moral reasoning is not law-like, and that when we are faced with forming our moral attitudes or acting morally, natural law can offer no contentful recommendations beyond the immediate moment. Why, one might plausibly wonder, appeal to a conception of law, if the aim is to endorse a radical particularism about moral choice and action? To be sure, these claims about origins are only part of a larger and more complicated historical story. The purely intentionalist or internalist account of natural law and morality that originates in the early Stoa gives way, in Vander Waerdt’s view, to a more rigid, rule-based theory under various philosophical and social pressures.2 Indeed, by the time of Cicero, perhaps, and of Aquinas, certainly, and then in all accounts of natural law thereafter, the tradition transforms its notion of law from an intentionalist into an externalist one, and its focus shifts from the virtuous inner dispositions of the wise to the delineation of hierarchical sets of rules governing the exter- nal behavior of all men. In a deep sense, Vander Waerdt’s overall assess- ment of the natural law tradition is one of depressing decline, both in its substance and in the expectations it has of its audience. It begins as a theory that attempts to ground in nature the kinds of rational motivations and dispositions we plausibly appeal to in explaining individual moral actions and that arguably are crucial for explaining accounts of the development and perfection of virtue. It declines into a theory that eliminates concern for virtue per se and with it, the regulation of agents’ motivations and inten- tions. It thus settles for the mere stipulation of external behaviors and 2 The historical basis for this claim is set out with great power and learning in Vander Waerdt (1989), (1991), and (1994). the stoics and aquinas on virtue and natural law 37 thereby abandons the project of cultivating our virtuous political and social impulses.3 As a consequence, what it offers is merely an abstract and rigid code of external conduct that cannot but fail to capture the particular nuances of a moral life characterized by wisdom and virtue. Sadly, one might note, this seems an especially ironic outcome for the theory given its origins in the Stoics’ fundamental belief that virtue alone is what is valu- able; indeed, presumably one of the primary reasons they held virtue to be uniquely valuable is precisely because it requires the autonomous cultiva- tion of one’s inner psychological states which, in their view, are in no way dependent on the vagaries of externals. In all stories of decline and fall, of course, there can be a kind of concep- tual clarity and sense of historical inevitability that makes for compelling and exciting reading. But at the risk of playing Bishop Spadeworth to Vander Waerdt’s Gibbon, I want to argue that the truth about the natural law tradition is considerably more complicated and that the basic elements that sustain it are there from the beginning and continue to persist, certain- ly at least through the formulations of Aquinas. One of the questions I therefore will be asking is why the tradition of natural law was able to offer viable philosophical alternatives for so many for so long, and especially for those who took themselves to be concerned with virtue and its cultivation. Moreover, since I see neither strong conceptual shifts nor decline, I do not think that we need to be on the lookout for any good or bad emperors in the story. No Stoic, even those for whom we have more evidence of an interest in delineating law-like rules of behavior, ever gave up a commit- ment to the inner criteria of virtue. Nor, arguably, did Aquinas. According- ly, although such a Spadeworthian history of natural law may be more conceptually tangled and certainly less exciting, it perhaps will turn out to be more uplifting in the end, since, after all, it is a history depicting both unceasing virtue and the abiding empire of law. It might be useful to begin with Aquinas, not only because his account is more familiar, but also because Vander Waerdt’s understanding of Thom- istic theory, I believe, is to a certain extent shaping his reading of the entire development of the tradition. Of necessity, I must be aggressively schema- tic, but it seems to me that Vander Waerdt’s account relies on taking the so-called Treatise on Law (and so the most famous passages on natural law) in isolation from the rest of Aquinas’ discussion of ethics proper. This is important, since it leads him, I think, into a parallel mistake about the relation between natural law and virtue in the early Stoa. In any case, although it might often go unremarked, it is quite obvious that Aquinas’ 3 See Vander Waerdt p. 23 with the quotation from Leibniz (1706). 38 phillip mitsis discussion of law follows upon an extended account of ethical theory or more precisely, if you will, virtue ethics. He discusses moral goodness (questions 1–17), morally right actions (questions 18–21), and the states of human beings that are essential to moral goodness and the performance of morally right actions — these include the passions (questions 22–48), habits in general (questions 49–54), virtue (questions 55–70) and vice (questions 79–81). Aquinas’ discussion of ethics thus begins with an account of happi- ness and concludes with problems of virtue. In introducing the Treatise on Law, Aquinas says that having treated of the intrinsic principles of human action, i.e. those of ethics proper, he is now going to talk about external principles. But this does not mean, of course, that his views about intrinsic and external principles are unrelated. What he has in mind in the Treatise on Law is primarily real law of that sort that imposes external constraints on moral agents. He is at most tangentially interested in Kantian-style moral laws or what we might characterize as moral precepts that agents use to govern their moral conduct. Indeed, Aquinas says ‘Properly speaking, no one imposes a law on his own actions, since personal moral judgments about what one ought to do are not laws, strictly speaking.’ (q. 93.5)4 Yet, within the Treatise on Law, one can also find abundant evidence of the intentional requirements of Aquinas’ wider moral theory. He says, ‘A man can bring law to bear on those things which he can judge. But a man’s judgment cannot be about interior motivations.5 Nevertheless it is required for the perfection of virtue that a man be right (existat rectus) in both sorts of acts.’ (91.4). This requirement, not surprisingly, serves as an important element in Aquinas’ claim that there is a need for divine law. Such divine dictates about our inner states can be found, for instance, in the Sermon on the Mount in which we learn that it is not only wrong to kill, but also to hate, not only to steal, but to covet, etc. Clearly, the fact that Aquinas’ central moral theory gives such a promin- ent place to virtue is itself evidence that he takes the internal states of agents to be crucial for moral evaluation. A virtue, for Aquinas, just is a disposition of the agent toward the right ends. Choice (electio) is therefore essential to morally right actions. At the same time, however, one might ask how these features of Aquinas’ moral thinking, which can be character- ized as being straightforwardly intensionalist, are supposed to fit with the picture of externally imposed natural laws that we find in the Treatise on 4 For fuller discussion, see Kretzmann (1988a), (1988b). 5 I am indebted to Scott MacDonald for discussion of these passages. With him, I take Aquinas’ reference here to motibus ‘that are hidden’ to be making a distinction between internal motivations and external behavior. the stoics and aquinas on virtue and natural law 39 Law. The question is central for Aquinas’ account of natural law and also absolutely parallel, I would argue, to a key question facing the Stoics — how to connect their intentionalist moral psychology, theory of virtue, and account of moral evaluation with a set of divine and externally imposed natural requirements. In the case of Aquinas, the connection between these two elements seems to be roughly the following. He often says that morally right action is ‘action in accordance with right reason.’ This formula for Aquinas argu- ably includes both externalist and intentionalist considerations. It is extern- alist because action in accordance with right reason is just that action that divine reason and natural law dictate. But the reasoning, as Aquinas says, is in the agent performing the action, and the agent’s following right reason in this case is a manifestation of the agent’s soul — that is his virtue — disposing him to follow reason in general. However successful Aquinas’ attempt to bring internalist and externalist elements together in one account, it seems to me that the early Stoics face a parallel problem in trying to reconcile the demands of their intensionalist accounts of virtue and happiness with what are taken to be external commands of a divine and providential reason. The same holds for the later Stoic tradition and, indeed, for natural law theorists well into the early modern period. Indeed, at no time, until very late in the tradition, does natural law become a theory focused exclusively on following a set of externally imposed commands. By the same token, it would be a mistake, however, to treat the early Stoic view as a purely intentionalist one since it is quite clear that, for them, the laws of nature emanate from divine reason — and hence are externally imposed requirements. It seems to me that something goes wrong with Vander Waerdt’s claim that natural law for the early Stoa is solely about dispositions. Clearly, as for Aquinas, disposi- tions and the reasoning that is in moral agents are an important part of the story for the early Stoics. But we have equally compelling evidence that natural laws for the Stoics are just that — laws of nature that are not entire- ly reducible to a series of particularist manifestations of the inner disposi- tions of moral agents, however infallible. One long-standing problem in the interpretation of Stoic moral theory has been how one connects evidence from their theories of virtue and happiness to their claims about nature and the divine. If one focuses on the Stoics’ accounts of virtue and happiness, one can surely find abundant evid- ence for their emphasis on an agent’s inner dispositions.6 It is sometimes 6 Here I simplify, since even in Stoic texts that some have taken to give strictly inter- nalist accounts of the perfection of one’s internal states and virtues, there still remain difficult questions about the kinds of roles that externalist considerations play in 40 phillip mitsis tempting to suppose that this is the whole story about their moral theory, and to then infer, as I believe Vander Waerdt does, that everything of moral importance that can be said about the sage’s perfected rational disposition appears in texts that describe the workings of virtue. Once one makes this step, however, it is easier to slide into a view that takes natural law to be ‘dispositional’. But it seems to me that the Stoic view of moral action is in many ways no more ‘dispositional’ than that of Aquinas, since it is clear that the sage’s reasoning is also in accordance with what are viewed as externally imposed divine requirements. Given Aquinas’ extensive discussion of both law and virtue, we would be compelled, even without the kinds of explicit indications that he offers, to attempt to piece together theses two aspects of his theory. The frag- mentary nature of Stoic texts perhaps makes it easier to lose sight of these twin demands on action in accordance with reason. A famous passage from Diogenes Laertius, however, reports the following claim of Chrysippus — a claim that in many ways, I think, prefigures the Thomistic linkage between internalist and externalist requirements and between virtue, happiness, and natural law: Therefore, living in agreement with nature comes to be the end, which is in accordance with the nature of oneself and that of the whole, engaging in no activity forbidden by universal law, which is right reason pervading everything and identical to Zeus, who is the director of the administration of existing things. And the virtue of the happy man and his good flow of life are just this: always doing everything on the basis of concordance of each man’s guardian spirit with the will of the administrator of the whole. (D.L.7.88, trans. Long and Sedley, 63C) The Stoics, of course, are innocent of the kinds of voluntarist and rationalist debates that this picture of God lets itself in for by the time of Aquinas and they have a monistic moral psychology free from tensions between reason and the will. But when the Stoic sage is said to act in accordance with ‘right reason’ or to be following the will of Zeus, this is meant by the Stoics to capture, in much the same way as it is in Aquinas’ theory, both internalist and externalist criteria. Vander Waerdt is surely right to argue that the cor- rect moral action of a sage mandated by nature’s divine reason expresses the sage’s internal rational disposition arising from his virtuous state of soul. There is little doubt as well that, to borrow a turn of phrase from Aquinas, the reasoning is in the agent performing the action. But the sage’s action is also in accordance with right reason and is of the sort that right reason and divine law dictate. Thus, to claim as Vander Waerdt does, that structuring agents’ reasoning and impulses toward their own good. See, e.g. Frede (1999) pp. 71–94. the stoics and aquinas on virtue and natural law 41 the early Stoic theory of natural law is dispositional is only part of the story. The full story includes externalist demands as well as an account of the relation of those demands to individual virtuous dispositions. It might be objected at this point, of course, that whatever the origin and nature of the commands of right reason, at least the actual content of its demands remains strictly dispositional, in the sense that nature’s com- mands for the early Stoics are about internal states of agents rather than about particular actions or external behaviors. I take it that this is an important linchpin in Vander Waerdt’s claims about Stoic particularism and, as far as I can make out, one reason for this is because he takes there to be a fairly straightforward connection between dispositional accounts and moral particularism. But, of course, there may be other routes to moral particularism. One might claim, for instance, that natural law merely gives a series of particular injunctions about moral actions (in the sense of external behaviors) on an ongoing basis. In other words, the claim that we can rely on no general rules to guarantee that we act infallibly does not mean we can only appeal to dispositions to guarantee such infallibility. Conversely, appealing to dispositions to explain moral actions does not necessarily require a commitment to particularism. One might argue that moral dispo- sitions or the reasoning behind such dispositions are susceptible of general analysis and thus exhibit law-like features that we can use to predict the types of actions to which they give rise. Thus, one cannot merely assume a straightforward connection between moral particularism and an interest in virtuous dispositions. Note as well, that for both Vander Waerdt and Inwood, the commands of natural law cannot be about the kinds of gener- alized dispositions we find, say, in Aquinas, not to hate or not to covet.7 They must turn out to be commands aimed at getting an individual in particular circumstances into a discrete psychic state that is causally linked to a particular action. It is not clear to me what philosophical sense, if any, can be made of this claim, but if we are to attribute it to the early Stoics and connect their interests in moral dispositions to any specific theory of particularism, it seems to me that we at least need some textual warrant. But here, the dispositional theory of natural law runs into difficulty. Vander Waerdt’s treatment of the evidence from Chrysippus’ Peri Nomou well illustrates our problem, since in his account8 there is not only, I would argue, a troubling elision of externally imposed divine commands, but also an assumption that the actual content of those commands is necessarily dispositional. At the same time, it seems to me that the evidence 7 Inwood (1999), for instance, argues that a general injunction to, say, act prudently, would be vacuous. 8 This volume, pp. 24–31. 42 phillip mitsis is neutral, to say the least, with respect to questions of moral particularism generally.9 Let us begin with the exordium of Chrysippus’ Peri Nomou. Law is king of all things divine and human. It must preside over what is honorable and base, both as ruler and as guide, and in virtue of this it must be the standard (kanon) of justice and injustice, prescribing to animals whose nature is political what they should do, and prohibiting what they should not do. (Marcian Inst.I = Long and Sedley 67R, adapted) The claim that Chrysippus formulates a ‘dispositional’ theory of natural law strikes me as problematic for several reasons. In this initial passage, which is admittedly programmatic, there is nothing to give a hint that the content of the injunctions of natural law is either about dispositions or, for that matter, about moral particulars. The early Stoics, just as Aristotle before them, have a well-articulated distinction between actions and their underlying intentions. Indeed, their ethical theory rather relentlessly stresses, some might say to point of implausibility, the claim that the only thing that matters from the point of view of moral evaluation is the intention from which an action is performed. But these worries about what, following Aquinas, we might call the intrinsic principles of human action are hardly in evidence here. Chrysippus does not claim that natural law prescribes to animals whose nature is political how and how not they should perform actions or with what sorts of inner attitudes. He maintains that natural law prescribes what they should and should not do. Given the centrality of intentions in their theory of virtue and also their fondness for invoking their own technical distinctions and vocabulary even in the most programmatic contexts, we should be wary of assuming the kind of neat folding of these externalist concerns into the requirements of ethics proper as postulated by Vander Waerdt. This is not to claim, of course, that Chry- sippus’ view of natural law is unconnected to his views of moral action. Far from it, since it would be a foolhardy to venture such a claim about a school that prides itself on the systematic unity of all of its philosophical positions. But the fact that particular Stoic views are connected does not mean that we cannot discern different philosophical motivations from separate areas of their thought coming into contact. Nor does it mean that we should merely run these separate motivations together or elide them. Of course, the question of how smoothly such motivations fit together is another matter. But at least at first glance, it seems implausible to suggest that this passage is concerned with the intrinsic principles of human action and not with the workings of an externally imposed law. 9 This volume, p. 28. the stoics and aquinas on virtue and natural law 43 Moreover, Chrysippus’ claim that natural law prescribes what we should do, is at best ambiguous between more general recommendations and those aimed solely at particular occasions. I say at best ambiguous, since it seems to me that only someone in the grips of a particularist theory could take this passage to preclude general law-like recommendations. But since such an interpretation cannot be entirely ruled out at this point, we need to turn to what is perhaps the crucial move of the dispositionalist interpretation. In claiming that natural law enjoins only dispositions, Vander Waerdt invokes the Stoic distinction between katorthomata and kathekonta, which is meant to pick out the crucial difference between actions that arise from the proper moral intentions and those that may not. He further claims that natural law enjoins only katorthomata. The sole passage, or better, sentence he relies on for this claim comes from a passage in Plutarch ( De Stoic. Repug. 1037c–d=SVF 3.520) whose aim is meant to show inconsistencies in the Stoic position. Here is Vander Waerdt’s reconstruction of the passage in question: They say: ‘Virtuous action is the prescription of law; moral error is its proscription. Hence law proscribes many things for the base, but prescribes nothing, for they are incapable of virtuous action. How can natural law prescribe virtuous actions for all animals that are political, Plutarch objects, since the vicious are incapable of virtuous action? Plutarch, I take it, is concerned with what becomes a standard objection to natural law theory, and for which there is an equally standard response. Natural moral laws are unlike physical laws and therefore can be dis- obeyed. Thus, their prescriptions are equally for the vicious as well as for the virtuous, even if the former do not obey them. However, even if we grant to Vander Waerdt that the passage reflects a Chrysippean claim that natural law enjoins only katorthomata (which I think is deeply problematic on textual grounds10 and also threatens to make kathekonta somehow anomalously natural), it still fails to show that the Stoic account is narrowly dispositionalist in the way that he suggests. Katorthomata are actions that are undertaken with the right rational intention and have the right focus on the good. Yet, even if natural law were held to enjoin only katorthomata, the most that this would show is that natural law for the early Stoics enjoins actions that are performed out of the right moral intentions. It does not follow that the injunctions of natural law themselves are solely about inner states or the dispositions producing 10 Since I have discussed this passage at length elsewhere (Mitsis, 1994), I will refrain from repeating my reasons for rejecting it as good evidence for the Stoic position. 44 phillip mitsis those perfect actions. If I return a deposit with the right moral intention and my action is in accord with right reason, I have performed a perfect action (katorthoma). But I have not merely lit up a particular inner bulb. I have performed an action and my action also has an external dimension. This intuitively plausible feature of morally correct actions is reflected in the Stoic claim that all katorthomata are also kathekonta. Part of the reason Vander Waerdt goes astray here, perhaps, is because of a related set of issues concerned with the Stoic response to the external failure of particular actions. Suppose, for instance, that I intend to return a deposit, and it is the morally right thing to do, but I am prevented from doing so by something out of my control. What is the moral status of my action? For the Stoics, such an action, even if it fails in achieving its external aim, is still entirely successful from the moral point of view. Accordingly, one might be tempted to conclude from this Stoic doctrine that successful moral actions can only be identified by focusing on their underlying intentions. Let us grant, for the moment, Vander Waerdt’s claim that the infallible sage is the paradigmatic moral agent acting in accordance with nature’s commands and that natural law only gives injunctions for success- ful moral actions (katorthomata). This still does not get us to his conclusion that katorthomata are internal dispositional states, however, or that the con- tent of nature’s commands is entirely dispositional. If natural law enjoins that a sage return a deposit and this is a perfect action (katorthoma), he might manage to return the deposit or he might be prevented from doing so by events out of his control. His injunction from nature, however, if it is in the form of an injunction to perform a katorthoma, is to perform an action. The fact that the external object of the action is attained or not is strictly a matter of indifference. But this further feature of our evaluation of the action does not show that the action itself is not part of what is enjoined by nature, however successful or unsuccessful it may be in attaining its external aim. This is because katorthomata are not merely dispositions; they are actions carried out from the right moral motives. Attention to the inner motivation of an action and indifference to an action’s external success does not commit the Stoics to the claim that actions themselves are merely dispostions to act. Thus, a sage obeys a moral injunction to act, not merely to get himself into a particular moral mood. Of course, it might seem odd, in the light of these claims, to say that whether the sage manages to return a deposit or not, he is performing the very same moral action (katorthoma). But the oddity stems here, not from the Stoics inability to distinguish dispositions from actions, but from the extremity of the moral perspective they espouse and their belief that success or failure in attaining the external aim of actions does not serve to individuate those actions, at least from a