Northeast Conference on Chinese Thought November 12–13, 2022 New Haven, CT How Far Is Xunzi from the Western Natural Law Tradition? Anderson Weekes New York [anderson.weekes@gmail.com] Abstract Interpretations of Xunzi’s meta-ethics run the gamut from relativist and constructivist to absolutist and realist. But, as David Wong argues in “Xunzi’s Metaethics” (2016), the fact that each of these readings can find its share of textual support in the Xunzi suggests that none of them is right. Xunzi is occupying some kind of middle position that eludes our dichotomies. If he’s an absolutist or realist, his position is softened by a recognition that the same standard, to be applied appropriately, must be applied differently in different circumstances. If he’s a relativist or constructivist, his position is qualified by recognition of objectively real limits on what will work. Most important is Wong’s clear-headed conclusion that a constrained relativism and a flexible absolutism are tantamount to the same thing. It’s not necessarily a contradiction if Xunzi is both a “relativist” and an “absolutist.” Taking a step back, I show that Xunzi’s peculiar ambivalence about meta-ethics is far from the exception. In general, meta-ethical descriptors tell us more about a philosopher’s starting point—her priorities in structuring and presenting the argument—than the position she ends up with. This is strikingly evident in western natural law theory. Although we tend to associate “natural law” with inflexible absolutism and extreme realism in meta-ethics, the tradition is not monolithic. Many theorists avow relativism and constructivism from the outset. But we need only read with the same degree of circumspection Wong affords the Xunzi to see that no theorist, regardless of persuasion, is consistent. Clear meta-ethical commitments one way or the other are tendered programmatically only to be hedged to the point of reversal in the course of execution. I document this by looking at a diverse sample of natural law’s most famous representatives (Protagoras, Plato, Aristotle, Aquinas, Ockham, Suarez, Hobbes, and Burke). In each case, we encounter the same patterns of ambivalence we find in Xunzi. I propose that meta-ethical ambivalence is an unavoidable consequence of the kind of approach to ethics Xunzi shares with natural law theorists. According to this approach, social norms are the solution to a design problem (the challenge of living together in order to prosper as human beings). The solution is found through some kind of actual or intellectually construed optimization procedure. But no solution can avoid being paradoxical because it must satisfy conflicting demands. To be relevant, it must conform to the specifics of actual situations, making it ad hoc. At the same time, it can’t be arbitrary. To be normatively valid, it must generalize unproblematically to all cases of similar type. But that’s possible only for a solution whose validity is objective and absolute. As a political philosopher, Xunzi was doing nothing unusual to characterize norms in seemingly opposite ways (constructed and objectively real; relative and absolute) or to think a correct account of norms would have to combine all of them. The comparison with individual western figures shows four different ways the combined position achieves coherence while meeting the conflicting demands of applied normative theory. The Problem with Xunzi One reason the Xunzi1 is so interesting to us is that it engages meta-ethical topics. What norms are, how they operate, where they come from, and what justifies them—these are all questions Xunzi poses. The problem, as any attentive reader can see, is that he does not appear to give consistent answers. 1 The ancient collection of Chinese writings known as the Xunzi is so called because it is traditionally attributed to the 3rd century BCE Confucian philosopher Xunzi, or Master Xun (荀子). While the collection as we know it cannot be the product of the master’s hand, there is no question that its distinctive doctrines derive from the celebrated teacher who studied at and 2 In “Xunzi’s Metaethics” (2016), David Wong has documented the extent of these vacillations by combing the text for answers to two cardinal questions.2 In regard to social norms, was Xunzi: 1) an absolutist or a relativist? 2) a realist or a constructivist? Wong’s analysis shows the textually justified answer is: Xunzi was all of the above. These tensions are clearly on display in the famous ch. 23 of the Xunzi (性惡, “Human Nature is Bad”):  Norms are not natural because they are produced by human effort and specifically by the almost supernatural perspicuity of the ancient sages (H 248–254; K 3.150–158).3 That means norms are constructed and their value relative to the needs that motivated us to construct them.  But, at the same time, norms are not artificial because anyone could become a sage just by employing their natural faculties correctly (H 254; K 3.158–9). That must mean norms are objectively real and absolutely right. How can Xunzi contradict himself in this way? Wong suggests three reasons for this state of affairs. A) We can’t expect Xunzi to give clear-cut answers to our questions—questions he himself was not asking. B) He was engaged in a synthetic project of reconciliation, mediating between opposing positions in the debates of his day. C) Our analytic dichotomies do not do justice to the complexity of some ethical positions. As for (A), I’m skeptical. Certainly the vocabulary of classical Chinese philosophy is different from ours, but I’m not sure the questions were so different. Whether norms are constructed or objectively real was very much discussed in pre-Qin philosophy. The question was broached in terms of the appropriate explanatory model for social development: manufacture vs. agriculture. Manufacture violates the nature of its raw materials and creates something new, whereas agriculture facilitates what already occurs naturally, removing the obstacles that would otherwise prevent something in its raw state from developing sua sponte. If the emergence of society is like manufacturing, then norms are constructed. If the emergence of society is like the cultivation of grain, then norms are objectively real. This is clearly a debate that was going on in the Warring States period (something Michael Puett’s book The Ambivalence of Creation makes abundantly clear4). Mengzi leans to the agricultural model and Xunzi to the manufacturing model. But both seem to sense that what they really need is a way to combine both models into one. Whether through blissful ignorance or profound insight, Kongzi himself was committed later headed the Jixia Academy in Qi. For an account of his life and the genesis of this remarkable text, see the Introduction to John Knoblock’s Xunzi: A Translation and Study of the Complete Works (Stanford: Stanford University Press) vol. 1, 1988. David Wong, “Xunzi’s Metaethics” (2016). In: E. Hutton (ed.), The Dao Companion to the Philosophy of Xunzi (Dordrecht: Springer Science+Business Media B.V.), 2016; pp. 138–163. 2 3 I reference the one-volume translation of Eric Hutton (Xunzi: The Complete Text, Princeton, 2014) by H + page number and the three-volume translation of Knoblock (Xunzi: A Translation and Study of the Complete Works, Stanford, 1988–94) by K + volume and page number. However, unless otherwise indicated, translations are mine. Because this paper is aimed at a wider audience than just professional sinologists, I have used the most easily accessible text of the Xunzi—the digital text made available on the Internet by the Chinese Text Project (https://ctext.org/xunzi). 4 Michael Puett, The Ambivalence of Creation (Stanford: Stanford University Press), 2001. 3 to both models, as is clear from a comparison of Lunyu 1.2 (the Way grows naturally from a secure root) and 1.15 (progress on the Way requires the moral equivalent of cutting, carving, grinding, and polishing). Something can be said on my reading for Wong’s option (B). Xunzi’s project is an attempted synthesis of opposites. But I think (B) is of vanishing relevance compared to (C). I’m here to radicalize Wong’s thesis. If we are taking western moral philosophy as our comparator, the kind of have-it-both-ways contradictoriness we see in Xunzi is not the exception, but the rule. It’s not just occasional synthesizers and middle-way seekers to whom our dichotomies do no justice. Rather, I claim: Any account of normatively-guided behavior fleshed out enough to apply to the complexity of a real situation will yield contradictory answers to Wong’s cardinal questions. I’d probably need several lifetimes to make good on such a claim. For now, I’m prepared to make a small (but not insignificant) down payment. I can show that “having it both ways” is the life and soul of what we know in the west as natural law theory and that Xunzi contradicts himself on Wong’s cardinal questions because natural law theory is exactly what Xunzi’s doing. I begin with a plainspoken account of what natural law is. I then summarize the widely-accepted view that very different paradigms divide the history of natural law theory into two periods, a pre-modern and a modern. What I question is not whether there are different types of normative theory on display in the history of natural law, but whether these types differ in the philosophically profound ways usually alleged. Since my workaday definition of natural law is crystallized from what the recognized varieties quite obviously have in common, we have a starting-point that everyone can accept. I then proceed to the purely theoretical part of this paper. I describe the operation of practical reason in a way that is unconventional, but undeniably accurate. Analyzing the implications of that description, I show that normative theory will indeed always take one of two basic forms, but, regardless which form it takes, will always end by approximating the other form. I turn next to documentation. Again and again I show how far western natural law theory is from its conventional representations. We’ll see that the history of natural law delivers exactly what my theoretical analysis leads us to expect. After summarizing the typology of natural law that the evidence from western sources has corroborated, I turn to Xunzi. I document his vacillations and show that, in every case, he is replicating a pattern we found in the western natural law tradition. I conclude with a resume of my speculations on the nature of practical reason and the constraints it imposes on normative theory. What Is Natural Law? “Natural Law” is the idea that some norms are self-enforcing. Self-enforcing doesn’t mean selfimposing. Non-compliance is possible, but it’s self-punishing. Everything that falls under the rubric of “imprudent” (incest, overindulgence, ingratitude) is an example. It naturally results in consequences the agent regrets. Despite its reputation, natural law theory is nothing grandiose. It builds on the mundane fact that specific effects (such as good health or social harmony) have specific causes. Success in achieving the desired effect depends on fulfilling the conditions that make it possible. Natural laws are the conditions of human functioning that we can identify by thinking about or being acquainted with the way things 4 are. Applied in a juridical context, they become the non-arbitrary reasons we can give for enacting positive laws in the first place or for critically evaluating them after their enactment. Recognized Varieties of Natural Law Theory When its concern is the functioning of the individual, we know natural law theory under the more familiar name “virtue ethics.” When its concern is the functioning of society, it takes on a deceptive variety of forms. Modern theories tend to focus on the minimal conditions of social functioning (keeping individuals unharmed), ancient and medieval theories on the optimal conditions (making members of the community virtuous or happy). Some theories are rationalistic (e.g., Hobbes), others empiricist (e.g., Burke), although most accept both ways of coming to know natural laws (e.g., Aquinas, Grotius). Natural law’s philosophical flourishes are famous, but comparatively rare. Typically, natural law argument focusses on the evaluation of conduct in specific cases, real or hypothetical. It usually proceeds by reasoning from or to general rules: either applying accepted rules to determine what conduct would be right in specific situations, or inferring what rules known conduct implies to determine if it agrees with the accepted rules. But it can take other forms. The Rationalism of the 17th and 18th centuries mistrusted the ad hoc character of traditional natural law methodology. It modeled the “compositive” part of its famous resolutive–compositive method on the constructive proofs of Euclidean geometry, allowing its more exuberant practitioners (e.g., Hobbes, Pufendorf) to make the attention-grabbing claim that natural law was an a priori science. PRE-MODERN TRADITION OF NATURAL LAW. The most important event in the history of natural law is the early modern rejection of Aristotle’s natural teleology. According to the pre-modern tradition, it is the natural state of a human being to be part of a community that is politically organized. Consequently, however much convention may contribute to the form a polity takes, fundamentally it is something that exists by nature. By the same token, no individual is answerable to herself alone because individuals are, in their essence, parts of a larger whole. Natural laws spell out the demands that the smooth functioning of the whole places on the behavior of its parts. It’s not hard to see how it would be in the interest of the individual to subordinate her immediate interests to the needs of the community. Failing to do so undermines the conditions of her own well-being. However, human beings can be ignorant, and their desires can, for various reasons, be stubbornly misdirected. Although natural laws are self-enforcing, the short-sightedness and corruptibility of human beings supposedly poses such a risk to social well-being that, to benefit from natural laws, we have to legislate them positively and enforce them coercively. MODERN TRADITION OF NATURAL LAW. Not all pre-modern philosophers can be classed as part of the mainstream pre-modern tradition just described. The ancient contract theorists Protagoras and Epicurus were outliers who became the inspiration for a new tradition of political philosophy in the 17th century. According to the modern tradition, it is not the natural state of a human being to be part of a community that is politically organized. Polity is an artificial construct outside of which an individual is answerable to herself alone. In a face-off reminiscent of the disagreement between Mengzi and Xunzi, what the pre-modern tradition called “corruption of human nature,” the modern tradition just called “human nature.” That means social coexistence—however naturally desirable it may be—is only possible under artificial constraints. For the moderns, natural laws are stratagems whose artifice must take the forms dictated by the nature of the challenge and the nature of the outcome sought. Simply put: the desideratum of peaceful polity presents a design problem; whether we discover them empirically or through rational analysis, natural laws are the design solutions that will work. 5 SIMILARITIES AND DIFFERENCES. The differences between modern and pre-modern natural law theory are important and have rightly been the focus of much scholarship. But we misunderstand both periods if we ignore what they have in common. Despite all their differences, mainstream theorists of both modern and pre-modern natural law agree on the fundamental point that polity has a purpose. Disagreements over principles (e.g., what constitutes pristine human nature and what corrupts it) and goals (e.g., what constitutes human well-being), or over the correct method for relating principles to goals, do not compromise the fundamental agreement that the purpose of polity is to facilitate and promote human well-being. For either group, natural laws are the objectively necessary conditions according to which that goal can be reached. The significance of this point is often missed. It can’t be stressed enough: Hobbes and his modern followers did not reject a teleological concept of the polity. They rejected the idea that the teleology was natural. Political theory was now seen as an art that seeks to achieve an unnatural state of affairs (peace). But for Hobbes no less than Aristotle norms were to be justified as the best or proper means to achieve a sanctioned end. What really distinguishes modern from pre-modern theorists is not whether they embrace teleology, but whether it’s a utilitarian teleology of desire or a metaphysical teleology of pre-established ends. In the coming pages I will argue that even this difference is not quite what it seems to be. While it seems to mark out fundamentally different meta-ethical positions, it really amounts to different ways of getting the same ambivalent outcome. If you begin with normative theory, as pre-modern theorists tend to do, ends seem pre-given and reaching them seems obligatory, but then you still have to deal with the practical problems of implementation—and you have to shift meta-ethical gears to do it. If you begin with implemented practices, as modern theorists tend to do, it’s not hard to see what strategies succeed at satisfying desires, but then you still have to deal with the theoretical problem of normative justification—and you have to shift meta-ethical gears to do it. The difference between the two approaches is real. Depending on which component is given methodological priority or seen as the foundation of the other, entirely different theoretical problems (specific application vs. generalized justification) threaten the project’s feasibility and become the focus of attention. But that, I claim, is only because which problem remains depends on which solution is front-loaded. In the end, both approaches seek an accommodation between the same two theoretical tasks. Of course, neither modern nor pre-modern theorists form homogenous groups. When I referred above to the modern and pre-modern “traditions,” I meant the kind of natural law theory that we think of as typical and mainstream for its period. Without question, realism and absolutism are prominent in the pre-modern period, constructivism and relativism in the modern. But there are important outliers in both periods. We create a dramatic caesura in the early 17th century only by ignoring the outliers and representing each period as less diverse than it really is. Once we see that, the larger pattern becomes clear: natural law theorists in all periods begin by taking different sides on which component of normative theory should get priority, but, regardless of which side they take, eventually compromise and resort to the methods of the other side. I will give abundant documentation of the pattern I’m alleging in the “Exhibits” section. The next three sections will explain it theoretically. I will argue that there is an ineradicable tension between the two constraints normative theories must satisfy. We classify normative theories based on which constraint they give priority, but every normative theory must satisfy both constraints: it must provide a basis for the justification of norms and it must fit them to the real world in which justification is sought. 6 Whichever constraint a theory satisfies first creates a problem for the other. Consequently, no natural law theory completes its task before walking back the principles on which it was declared. A Parable About Practical Reason We can learn something about natural law from that arch-nemesis of natural law theory, Jeremy Bentham. In his Principles of Morals and Legislation, Bentham distinguishes four kinds of sanctions for misdeeds: religious, moral, political, and physical. Of the last he says: A suffering which befalls a man in the natural and spontaneous course of things, shall be styled … a calamity; in which case, if it be supposed to befall him through any imprudence of his, it may be styled a punishment issuing from the physical sanction…. A man’s goods, or his person, are consumed by fire. If this happened to him by what is called an accident, it was a calamity: if by reason of his own imprudence (for instance, from his neglecting to put his candle out) it may be styled a punishment of the physical sanction… (PML, ch. 3) After noticing that the most vehement opponent of natural law theory (it’s “nonsense on stilts”) is acknowledging the existence of natural laws in the exact sense understood by the natural law tradition, I want to ask what basis there is for Bentham’s distinction between “calamity” and “punishment of the physical sanction.” Jack and Jill are farmers. The Almanac predicts a one-year drought occurs every ten years. In order to be prudent, Jack saves enough to survive one drought year. But Jill thinks weather is not entirely predictable. Doesn’t prudence counsel being prepared for the unexpected? She saves enough to survive two years’ drought. Now a two-year drought occurs. Some will say that Jack was prudent, but suffered a tragic setback due to events he could not have anticipated. About Jill they will say she was lucky because, as it happened, she was overcautious. But others will say that Jill was the paragon of prudence. Jack, by contrast, was not unlucky, but foolish. We can draw a swarm of morals from this parable. 1. Prudence is what we call it when we manage to invent more or less reliable strategies for dealing with the vicissitudes of chance. But when expectations fail to anticipate the vagaries of life, whether we call it calamity or natural sanction for imprudence is largely arbitrary and depends on just how “examined” one thinks the examined life should be. Let this be our first moral lesson. 2. We can assume that Jack learns his lesson. For those who called his situation calamity, he has learned to be prepared for a possibility of pure chance. His adaptive response to bad luck appears to be something invented on the fly to meet needs that came from nowhere (read: it was constructed and relative). For those calling his situation imprudence, he has learned about the conditions of success inherent in the nature of farming, and he has learned the hard way what he could have known in advance through a rational analysis of risk factors in farming. Those who credit themselves with a more rational approach than Jack justify their farming practices on the grounds that the nature of the challenge places the same conditions of success on all farmers (read: their operational norms are objective and absolute). The point I’m presently drawing out of this story (call it Moral 2) is one part of the main thesis the rest of this paper is intended to sustain. Just as calamity and imprudence are two ways of describing the same state of affairs (per Moral 1), so, too, are social theories that see norms as constructed and relative and social theories that see them as objective and absolute. For convenience, I will refer 7 to the former as the utilitarian approach and the latter as the absolutist approach. They differ in what they choose to advertise and emphasize, not in what they finally acknowledge to be the case. 3. The advance risk analysis of the rationalists has the important characteristic of being principled (“a priori”) only after the fact. After someone has experienced a first-of-its-kind calamity we seek an explanation. Once we find it, we see how the calamity could have been rationally anticipated and therefore prudently avoided. We will know that for ever after. But could we have known it in advance?  Yes: the correct explanation in terms of the nature of things was already true.  No: we didn’t even know there was something we didn’t know. Before the unhappy consequences of our ignorance present themselves adventitiously, we have no idea what needs explaining, never mind anticipating. All ethical thinking—whether “philosophical” or not—operates in the shadow of this antinomy. That it is an antinomy will become clear, if not from the logical considerations of the next two sections, then by induction from the uniform pattern evinced in the otherwise widely varied exhibits that follow. We will see that normative social theory always succumbs to the same contradiction, which I’m calling the antinomy of practical reason. This point (call it Moral 3 of the Jack and Jill story) completes the main thesis of this paper. Utilitarian and absolutist approaches are not the opposites they are taken to be. They both confront the same contradiction. They confront it because it’s embedded in the very attempt to do normative theory. Consequently, the difference between them is not how they resolve the it, but, given the differences in their starting points, what they have to do to make their peace with it. Philosophical Implications of the Parable: Why Natural Law Theory Always Seems to Contradict Itself My parable about practical reason shows that the norms of rationally justified behavior are always both constructed and objective, both relative and absolute. Here’s the philosophical analysis: All behavior is contingently motivated. The point of rationally justifying behavior is to eliminate the contingency, replacing reasons of the first type with that of the second:   First Type: I did x because someone told me to; I felt like it; I didn’t know what else to do. Second Type: I did x because it was the right or best thing for anyone in my situation to do in precisely those circumstances. Prudence is behavior equipped with reasons of the second type. But contingency can’t really be eliminated. Prudence would be a fantasy about possible worlds, not the solution to real problems, if the challenge did not present adventitiously, e.g., in the form of what Bentham calls “calamity.” On the other hand, our ultimate policy would not be rational if analysis did not succeed, albeit only in retrospect, at making what happened “predictable” and “avoidable,” rendering the term “calamity” inappropriate. What does this have to do with natural law? Notice that justification of the second type is sub specie aeternitatis. To be sure, if the behavior in question is real and not just imagined (say, by a novelist), then it’s contingently motivated. To put it another way: it actually made sense in a real historical circumstance. However, the speaker is justifying it by appeal to an “eternal” rule—something that’s not historically contingent. And there it is: the thing that seems most to distinguish natural law theory from 8 everyday reasoning—the appeal to “eternal laws”—is no difference at all. Far from being the relic of medieval theology it’s often taken to be, it’s a factory-installed part of all rational justification. It’s no more “metaphysically freighted” than everyday attempts to rationally evaluate behavior. By the same token, however, everyday attempts at rational justification are no less metaphysically freighted than natural law theory. Making a non-arbitrary judgment about a real situation forces us to think of it as a “type” of situation, rather than the contingent token it is. It’s the type that allows us to engage in modal reasoning and draw conclusions about what is “necessarily” or “always” right or wrong. But now we come to the perennial trouble in rationalist paradise. The all-important type interests us only because the token does. Justification may seem like a purely theoretical affair. But the need for justification doesn’t begin with theory. It begins with anxiety about the outcome of a real situation we are confronting. We shift to thinking about the type because the best response is the rationally most defensible one, which we determine based on our knowledge and experience of the type of situation that confronts us. So a kind of ontological split emerges. We think about what type of situation it is, but only because this situation is “really happening.” The problem is the contingency of the situation is not expunged, but actually presupposed, by the effort to “rationalize” it (to classify it as this or that type and to analyze it as having such and such characteristics). The result is the same paradox we saw with Jack and Jill. The necessities of the type (the patterns prudence can intelligently anticipate) duplicate the contingencies of the token (the experiences of luck and calamity that no one could have anticipated before their first occurrence). An illustration will clarify the point. Readers of the early modern systems of natural law sometimes object that Pufendorf and his international cohort were misguided zealots who tried to absorb the whole of positive law into natural law by pretending to rationally construct it.5 But rationally constructing positive law is exactly what natural law is supposed to do. Positive laws are contingently motivated. Natural law theory tries to eliminate their contingency by validating them, in a rigorously controlled way, in light of accepted principles. For Scholastics, that meant seeing if positive laws could be deduced from assumed first principles by specifying empirical boundary conditions. For the systembuilding rationalists of the modern period it meant seeing if positive laws could be replicated by constructing them more geometrico from primitive postulates about human knowing, feeling, and doing. The “resolutive” part of their resolutive–compositive method supposedly isolated the simple elements of social order. The “compositive” part was then free to aim at reconstructing positive laws as we happen to known them. On either procedure, the driving question is which laws can be rationally validated (deduced or reconstructed) and which need to be improved or vacated. The starting-point is doubt about the validity of positive laws, which arises precisely because ultimate justification is never the motive for their enactment. But the end-goal is not to formulate purely rational laws in place of positive laws. It’s to evaluate (and, if necessary, modify) the laws we actually have. In other words, natural laws are nothing more than evaluation criteria. The principles are the ultimate criteria, and the The most informed representative of this group would have to be Heinrich Rommen. See his Natural Law: A Study in Legal and Social History and Philosophy (Indianapolis: The Liberty Fund), 1998; pp. 93ff. Rommen lauds Anselm Desing (1699–1772), “a Catholic, in contrast to the majority of natural-law teachers,” who pointed out “that the pretended natural law of his time was in no way a ‘dictate of reason’; that it was rather a rationalization of the positive law of the period, yes, even of the laws of the nation to which the author belonged; hence that it was not at all derived, as asserted, from reason alone, but was little more than ‘the civil law adorned with some spoils of philosophy and moral theology’” (p. 94). Rommen gives no citation but must be referring to Desing’s 1753 Jus Naturae liberatum ac repurgatum a principiis lubricis & multa confusione per doctores heterodoxos inductis (Natural Law, freed and purged of the slippery principles and manifold confusion introduced by heterodox teachers). 5 9 rationally-derived laws are the philosophically well-grounded reasons one can give for approving or disapproving of specific positive laws once they have, for whatever contingent reasons, been enacted. Notice how the process works. There is no parthenogenesis of rational precepts from natural law principles. First we must know what laws we hope to derive, then we can test them to see if they are rationally derivable. Without the input of a positive law that is, if not enacted, at least proposed in really-occurring historical circumstances, natural law can do nothing. Its role is corrective, not generative. So we have an excellent illustration of the antinomy exposed in my parable. Between ratifying laws that don’t need correction and correcting those that do, natural law theory aims at the very paradox Pufendorf is accused of: establishing a body of positive law that is completely rational. Contrary to what Pufendorf’s detractors assume, the implicit goal of natural law theory is a complete body of philosophically validated positive norms. The only thing that makes Pufendorf and his imitators stand out is how monumentally close they got to completing the task of validating (or correcting) all the positive norms of their day. But note well: the goal of natural law theory—even in its most rationalistic form—was not the invention of an entirely pure body of law derived from reason alone. The justifications were purely rational, but it was neither perversity nor coincidence that the constructions of the rationalists replicated as closely as possible the positive laws of their day. It’s this world we want to be right about. The view of natural law endorsed by Pufendorf’s critics is a very common one. It assumes that the process of justifying a law is also the process of discovering a law. But that’s wrong—even in cases where the natural law theorist corrects positive laws. Why? Because in order to be an adjustment of positive law, natural law must first aim at adjusting itself as much as possible to positive law. It’s only in cases where the adjustment to positive law proves impossible that the natural law jurist is justified in demanding the adjustment of particular positive laws. In fact, he has no other way of knowing what to demand. So the primary responsibility of the natural law jurist is to attempt the very thing Pufendorf is accused of: absorbing positive law into natural law. Are positive norms that are validated in this way (by systematic recourse to universal principles) absolute or relative? The answer is both. The point of legislation is to meet locally-recognized needs, but the point of justification is to universalize. As positive, laws are relative. As rationally validated, they are absolute. Are such laws constructed or objectively real? The answer is both. Legislation invents them as a response to contingent needs. There is, in fact, no other way for us to arrive at them. But then reason comes along and recognizes that many of these same laws can be justified as constraints of justice following from the nature of things. One objection to my argument might be that it’s tailored to a cherry-picked example. Famously, Pufendorf claims from the outset to be both a utilitarian and an absolutist.6 Should we be surprised that he ends up contradicting himself? But I’ve just explained how the Scholastics, avowedly single-minded in their absolutism, fit the same pattern as Pufendorf. A similar case can be made even for those vocally self-described utilitarians such as Bentham and Mill, who—because they see it as a proxy for the absolutism they abhor—claim to reject natural law altogether. Consider that “utility” and “fulfilling the natural purpose of society”—however different they are—are still different ways of talking about human See On the Duty of Man and Citizen, ed. James Tully (Cambridge: University Press), 1991; Bk. I, ch. 3 (“On Natural Law”), pp. 33–38. 6 10 flourishing. Utilitarians are less interested than absolutists in law as a logical system, focusing more directly on the consequences of the practices underlying positive laws. Indeed, that’s what makes them utilitarians: the consequences they care about are the ones we know about empirically because they actually occurred. Nevertheless, they care about those consequences only because they want to adopt the best policy going forward. But that means the goal of their legal theory is still the rational validation (or critique) of positive norms of behavior in light of a higher normative principle (utility). Rational validation for absolutists means seeing if norms enjoining particular practices can be derived from more general standards. For utilitarians, it means seeing if general standards can be derived from particular practices. It’s the same logical test, only approached from the opposite side. The all-important question for the utilitarian is whether practices that happened to satisfy people in some cases can reasonably be expected to do so as a rule (i.e., in most cases of some definable type of situation we can expect people to continue encountering).7 It’s still a matter of using rational argument to eliminate the contingency of something we wouldn’t care about and couldn’t possibly know about if it hadn’t arisen contingently in the first place. Of course the real question here is not whether we can find examples of normative theory that are, on close analysis, meta-ethically incoherent, but whether we can find any examples that are not. Since that’s far too wide a net for one paper, I have narrowed the focus to natural law theory, the long and varied history of which still leaves us with an extraordinarily rich sample to judge from. Isn’t there some version of natural law theory that escapes the antinomy I’m describing? The reason for this paper is my confidence that a sober and sustained look at the history of natural law theory will convince us otherwise. However, I want to be clear why, on theoretical grounds, my answer is already no. Rationally validating contingent practices is what normative theory does. Since that means providing a secure basis for the justification of norms and fitting them to the real, historically contingent world in which justification is sought, there’s no way of proceeding that won’t end with a coincidence of meta-ethical opposites. Absolutists can’t leave off talking about the actual world any more than utilitarians can refrain from counterfactual reasoning. Forestalling Some Obvious Objections to the Argument My analysis of normative thinking raises some obvious practical questions and invites at least one sweeping theoretical objection. Dealing with these problems now will forestall misunderstanding and give the historical analysis I’m gunning for a sharper focus. I tackle the theoretical objection first. “Explanation of facts by general covering laws” is a familiar notion in contemporary philosophy. I have highlighted how rational understanding always involves some form of it. I’m not suggesting that it always takes the complex form made famous by Carl Hempel. However, just seeing an individual thing, event, or relation as the token of a type is already a kind of explanation. It implies a set of predicates whose co-occurrence is not haphazard, but unified to the extent that it exhibits a recognized pattern. My point was that we can’t understand anything without generalizing in a way that turns the item in question into an instance. The item becomes intelligible precisely because we assign a repeatable identity to it. I then drew attention to how such explanation depends on and presupposes the very facts it seeks to explain. The objection we have to consider is that the dependence in question, while real, is trivial and doesn’t have any of the philosophically problematic consequences I allege. I say the contingency the token (the adventitious character of the experience that doing such-and-such in such7 Note that the debate between act and rule utilitarianism is a debate about higher-order rules. 11 and-such situation had a good or bad outcome) is never really eliminated by its rational explanation because theory wouldn’t know what states of affairs need to be rationalized if they weren’t given adventitiously in the first place. But is that anything more than a truism? Obviously it was the paths the planets actually follow that Kepler and Newton sought to explain with universal laws, but that doesn’t make the universal laws that explain them contingent! However, the two cases are not analogous. Unlike human beings, the planets don’t interact with one another and then change their paths in light of shifting goals. True, scientists are always making new discoveries about the behavior of nature, but nature doesn’t continually display new behaviors. By contrast, history presents the social theorist with new facts—not just newly discovered facts. In a fundamental way, normative social theory aims to generalize about the value of actions just as natural science generalizes about physical events: both seek a principled explanation for certain outcomes sub specie aeternitatis. But, however beholden natural science is to emerging empirical data, normative theory is nevertheless beholden to emerging empirical data in a way that natural science is not. It must continually revise itself not because long-existing facts are always being newly discovered, but because new facts are always emerging. This is the source of the antinomy of practical reason. The postulate of all explanation is that the contingency of the explanandum is only epistemic and disappears once the facts are explained. Social theory has to make this assumption, too. But in truth its explanandum possesses a contingency that is more than epistemic. That’s why historians and social scientists are very good at explaining, after the fact, why events happened the way they did, but unlike natural scientists, very bad at predicting anything that hasn’t happened yet. Just as history is “predictable,” but only in retrospect, normative theory is universal (“eternal” if you like), but only after the facts it explains have emerged. That’s the paradox, and it becomes acute precisely because new facts are continually emerging. Readers will certainly be justified in asking why, if I’m right that this antinomy is inescapable and ubiquitous, my thesis isn’t long-accepted conventional wisdom. But in a way it is: as unconventional as my thesis sounds, I am only assigning systematic philosophical significance to something I think everyone, at least implicitly, already knows. Even before I present my historical documentation, I challenge readers to think of an absolutist or a utilitarian theory that does not fit the patterns I’m alleging. A more penetrating question would be why we remain committed to standards based on the oppositional categories of meta-ethics if no one can think of cases where they strictly and properly apply. Every student of philosophy knows absolutists hone their principles by working backwards from the contingent world to which they must apply. But absolutists aren’t “cheating” when they come up with eternal principles perfectly fitting the social milieux they happen to be in. They are doing exactly what they signed on to do. Absolutism was never about insisting on pre-ordained rules, outcomes be damned. Theorists qualify as “absolutists” because they want to base ethics on entirely right rules, and they understand “entirely right” as “eternally right.” But all that means in practice is they begin with the conviction that an outcome is right only if it would result from following a generally acceptable rule. The problem—as every absolutist knows—is how that cuts both ways. A rule isn’t acceptable if it doesn’t produce a right outcome, and two familiar challenges make formulating an “always” right rule tricky. First, when previously stable (and therefore unnoticed) boundary conditions shift, the same rule can produce a different outcome. Second, conduct is subject to many rules. A rule is not entirely acceptable as formulated if, in any circumstance that happens to arise, it produces an outcome that is wrong in light of some other applicable rule. So there’s nothing surprising about it. Given the complex and evolving nature of the challenges, the business of absolutists has never been the pronouncement of eternal, 12 unchanging rules. It’s always been the continual revising and qualifying of what should count as the eternal and unchanging rules. On the other hand, every student of philosophy knows utilitarians aren’t interested in what works if they can’t make universalizing claims. True, they qualify as utilitarians because they base norms on what works. But all that means in practice is they begin by looking at what’s actually working. That’s where they start from, not where they stop. They never advocate for practices simply because they happened to succeed. They are interested in causation, not luck. They always test their claims with counterfactual scenarios, justifying types of practices on rational grounds. Just as the absolutist needs to be right about this world, the utilitarian needs to be right about it. Either way, the outcome is paradoxical. The absolutist claims adversity taught an eternally true lesson, but continues to revise that lesson with every change in circumstances. The utilitarian claims to reject eternal verities in practical matters, but won’t call a strategy useful unless reasons can be given why it would always succeed if the type of situation is the same. I’m not denying that absolutists and utilitarians differ on matters of principle. Absolutists think normative theory must be grounded on the justification of norms, utilitarians on real-world fit. But theorists of either stripe always proceed to recognize the other constraint as the most urgent problem their grounded theory has to solve. So they employ in practice what they opposed in principle. “Always” performs the same function for utilitarians that eternity does for absolutists. They need its counterfactual magic to turn the description of a successful practice into a normative theory—a nonarbitrary statement of what people should do. Of course, nothing stops utilitarians from being open about how provisional and hypothetical their “always” is, but that still doesn’t distinguish them from absolutists. Why? Because this is where absolutists take their turn emulating the opposition without admitting it. Revising our understanding of what the eternal rule is performs the same function for absolutists that changing the rule does for utilitarians. Stated in the broadest terms, then, my claim is that natural law theorists, despite occupying an extraordinarily long tradition that stretches from ancient Greece to the present, and despite occupying an extraordinarily contentious tradition that seems to be marked from the start by polarizing disagreements, in fact offer us remarkably similar theories. They are similar because, regardless of the position they claim to take on questions of meta-ethics, they always converge on the same “all of the above” position. I have explained why this is the outcome one should expect: normative theory can’t help being conflicted because its task—rationally justifying contingent practices—means negotiating between contradictory demands. Because you can start with either demand, normative theories will break down into two apparently opposed types. But normative theories must satisfy both demands, so both types eventually concede what they seemed to oppose. The remaining question is why we persist in wanting to characterize normative theories with the oppositions of meta-ethics if the categories never really fit. The fact is that theories come pre-packaged that way. Because you have to start with one demand or the other—and they really are in conflict— theorists are naturally tempted to think that choosing against the alternative is a foundational requirement of ethical theory. Virtually all philosophers present things that way, and we follow their lead. So when a theorist violates her declared principles, we grant the objective and query her inconsistency in carrying it out. But if all theorists violate their principles, a better strategy might be to question the feasibility of the objectives. 13 That natural law theories evince the patterns I allege is what the remainder of this paper will show. I will start with natural law theories in the Exhibits section and show how they fit the pattern. I then turn to Xunzi and show how Xunzi fits the pattern because he, too, is a natural law theorist. Transition from Conceptual Arguments to Historical Documentation The last two sections have delivered a clear and—I believe—compelling analysis of how practical reason operates. However unconventional my thesis may be (that the norms of rationally justified behavior are constructed and objective, relative and absolute), it follows directly from that analysis. But how does my thesis fare empirically? Pufendorf may be a compelling example of someone who is “all of the above”— certainly his critics think so. But I want extend the same analysis to all social theorists. Is Pufendorf as representative as I claim? I have challenged my readers to think of counterexamples. Is there a wellknown clutch of them that I must explain away in order to make my case? Surprisingly, the answer is no. Here, on empirical grounds, is where my case is strongest. The meta-ethical tensions I have teased out of my analysis of practical reason run conspicuously through the entire history of natural law theory, as the following exhibits will show. My theoretical analysis makes several claims: practical reason always contradicts itself; there are different ways this plays out; because they are different approaches to the same problem—one inherent in the nature of normative theory—the end-result is always the same. Since each of these claims needs to be corroborated, the exhibits have several objectives. The first is to demonstrate how common it is for western natural law theorists to answer our cardinal meta-ethical questions by checking “all of the above.” The second objective is to document that not all theorists have the same reasons for checking “all of the above” and to identify the most frequent rationales they have for doing so. The third objective is explanatory: to expose the antinomy of practical reason as the common denominator—the reason why different theorists, despite committing to very different projects, always end up with the same “all of the above” result. In the course the exhibits I uncover four distinct rationales for western philosophy’s “all of the above” ambivalence—one that is characteristic of those who advertise as absolutists and realists and three that are paradigmatically associated with one or more philosophers who advertise as constructivists or relativists. That the latter are not “double agents,” but legitimately belong to the same natural law tradition as the former is something the exhibits themselves will demonstrate. I do not claim that there are only four reasons for meta-ethical ambivalence in western philosophy. I claim there are at least four. There are probably others, but the consistent pattern of the exhibits leads us to expect that they will always be variations on one of the two basic approaches already identified, the absolutist and the utilitarian. Even if we bracket the analysis of prudence and calamity I gave above, it’s not hard to see why these two approaches might exhaust the field of possibilities. We engage in normative theory because motivation is complex, channeled both by a variety of general goals and by incidental conditions of success (if I’m hungry, I might go to a restaurant because it’s vegetarian or because it’s the only one that was open). Seeking coherence, normative theory either begins with general intuitions about what’s right and then asks how to make them work out in the face of real world conditions, or it begins with behavior that satisfies our desires in a real situation and then asks if it’s right (i.e., if, in general, that behavior would lead to approved outcomes). If the exhibits succeed in showing that natural law theories do typically take one of these two approaches and that both approaches straddle the same contradiction, then we have good reason to think my theoretical analysis is correct. The 14 implication is that every social theory will succumb in one of two basic ways to what I’m calling the antinomy of practical reason. One important question this paper answers is whether Xunzi’s ambivalence shares any of the same rationales as his western counterparts. The textual evidence I present later in this paper suggests he is no stranger to any of the four rationales I identify in the next section. This must count as not insignificant corroboration for the universality of my thesis. So that no school of natural law thinking goes unrepresented, I include in my sample those typically classed as arch-enemies in the absolutism–relativism debate (Plato–Protagoras, Aquinas–Ockham) and in the realism–constructivism debate (Aristotle–Hobbes), as well as the outliers on questions of methodology (Hobbes denies there is any empirical knowledge of natural laws; Burke denies there is any rational knowledge of them). We will see that they all follow the pattern my analysis leads us to expect. What differentiates them is not where they stand, but how they get there. Starting out from very different advertised commitments, some taking up positions on a platform of absolutism, others on a platform of utilitarianism, they are variously motivated to chip away at their own platforms until they meet at the same “all of the above” position in the middle. As complex as my thesis may be, the argument and methodology of the following section are straightforward. A pattern as counterintuitive and yet as robust as the one I document can hardly be coincidence: all natural law theorists contradict themselves, and committing to opposite meta-ethical positions just means they contradict themselves in complementary ways. I offer my analysis of practical reason as the best explanation. Exhibits: Hobbes, Aquinas, Aristotle, Suarez, Ockham, Burke, Plato, and Protagoras I am claiming that figures as famous for their differences as Hobbes and Plato understand norms in fundamentally similar ways. If that claim seems incredible, I can perhaps be forgiven for eschewing the stricture of chronology. I order my exhibits in a way that is most effective for documenting the similarities I allege. I begin with Hobbes because he can serve as a clear point of reference for the rest of the discussion. The tensions I want to draw attention to are, in his case, obvious and extreme. My exhibits fall into two groups: an initial series (Hobbes, Aquinas, Aristotle, Suarez, Ockham, and Burke) covering paradigmatic figures in both the modern and the pre-modern natural law traditions and a capstone comparing Plato, the recognized archetype of realism and absolutism, and Protagoras, the recognized archetype of constructivism and relativism. The figures in my first series of exhibits prove that neither modern nor pre-modern natural law theorists saw realism and constructivism, or absolutism and relativism, as mutually exclusive. Despite their very different starting points and motivating rationales, execution of their projects requires them all to engage in the same key manoeuver: an act of construction that presumes to disclose—not oppose—a reality that is mindindependent. They resort to this manoeuver for different reasons, but they all have to resort to it. I suggest this is why their projects fit neatly in our meta-ethical categories only in conception, never in execution. I complete my data set by comparing Plato and Protagoras. Conventionally, one would see each of them as embodying one of the two extremes we find combined with such explicit paradox in Hobbes. However, I show that each of them embodies both extremes as much as Hobbes himself does and that they do so not despite themselves, but knowingly and quite deliberately. They, too, embrace a rigorously controlled form of construction as the way for finite minds to discover what’s objectively real. 15 THOMAS HOBBES. The whole point of Hobbes’ political philosophy is that social norms do not exist by nature. What’s natural is a war of every individual against all others. Society is possible only if it is constructed. That’s why social norms must be created by an artificial contract.8 There could not be a stronger statement that norms are constructed. Yet Hobbes himself commandeers the language of natural law to describe his project. Why? He could have denounced the very idea of natural law as a relic of the Aristotelian natural teleology he abhorred, but he didn’t. Hobbes, in fact, does not deny that rational justification has to bottom out in eternal, non-contingent principles. He just thinks they have to be sought for in reason rather than in nature. He was satisfied that he achieved his goal. Because he thought he had proved a priori that only one solution to the problem of mutual mistrust was possible, he saw his solution as objectively discoverable and absolutely right—even though it had been humanly constructed (i.e., by a man called Thomas Hobbes). The entire early modern natural law tradition followed Hobbes on this point. Natural laws were natural because they were the rationally constructed solution to the unchanging problem our nature presents us with. However, it would be a mistake to think the important difference between modern and pre-modern natural law theory lies here. For ancient and medieval thinkers, natural laws are constructed—without cost to their objectivity—just as much as they are for Hobbes. The difference lies in whether the construction is seen as the fruition of a pre-existing natural tendency or as a de novo fabrication licensed by reason alone. A careful look at the “Treatise on Law” in Aquinas’ Summa Theologiae9 will bear out this claim. THOMAS AQUINAS. According to Aquinas, who is following a tradition that goes back to Cicero and Augustine, natural laws are supposed to be eternal and unchanging. However, in the universe of Catholic theology, nothing is eternal but God, and God is in no way subject to law. So how can there be eternal laws? Law is eternal and unchanging only insofar as it exists in God, and it exists there only as His eternal plan for creation—the way a design pre-exists in the mind of the engineer. However, to perform its regulative function, law must be applied and operate in time. Consequently, natural law is not strictly the same as eternal law. Eternal law takes two forms. In its original form, it eternally pre-exists in God, as His plan. Only in this form is it “eternal” in the true and proper sense. In a derived form, the same law operates in creatures, as God’s temporal government of creation. Natural law is identical to eternal law only in the derived form. It is not eternal law simpliciter, but eternal law insofar as it exists in the appropriately constrained and qualified way in the creature. What is that appropriately constrained and qualified way? That’s trickier than popular interpretations want to have it. It’s not necessarily wrong, for example, to think of natural law as God’s precepts and 8 See Hobbes, Leviathan, “The First Part, Of Man.” Summa Theologiae, Ia IIae, qq. 90–108. It is not uncommon to see the “Treatise on Law” cited (or even reprinted) as Questiones 90–97, as if the treatment of religious law (the “old” in qq. 98–105 and the “new” in qq. 106–108) was irrelevant. Given the scope of my topic, my treatment of Aquinas is necessarily quite abbreviated and draws only on qq. 90–97. However, an adequate comparison with Xunzi, for whom ritual is normatively paramount, would have to consider in some detail Aquinas’ treatment of the “ceremonial precepts” in the Old Law (Ia IIae, qq. 101–103) and the corresponding treatment in Pars Tertia of ritual in the New Law (“Treatise on Sacraments,” IIIa, qq. 60–90). For dogmatic reasons, Aquinas refuses to acknowledge that there is a ceremonial aspect to the sacraments, but he is happy to describe, analyze, and explain them (see q. 60) in a manner altogether analogous to his account of ceremony in the Old Law. One finds that for Aquinas, just as for Xunzi, ceremony has both functional and symbolic value. 9 16 commands, but that blurs the fundamental difference between natural law and divine positive law, creating a misleading picture of the how the different levels of law are related.10 Natural law is not eternal law issued as a command to the creature, like an afterthought or reminder. Eternal law already exists in the creature as the way the creature has been designed. It manifests as minima and maxima in the creature’s characteristic way of operating—as a set of outer limits on that way of operating and as an optimal range somewhere within those limits. But here’s the crucial thing. Operating in time means operating in response to contingent circumstances. God’s plan was not that the human being be able to function optimally in only one set circumstances—as if there was only one right value for every input, and the system begins to fail if all the inputs are not exactly right. Rather, the design is meant to function across a sprawling diversity of circumstances—which can only mean there is more than one optimum. It follows that natural law has different implications—different applications—in different circumstances. These different applications are nothing other than human (positive) laws insofar as they are just. Now let’s notice how peculiar the transition is from pure to applied law. In the final analysis, a human law is just only if it stands in some relation of agreement to eternal law. So how do we get from eternal law to human law that is in conformity with eternal law? If the transition were purely deductive and depended on no contingent inputs, we would not be talking about an application, but about an operational feature of the eternal design. On the other hand, if the eternal design did not constrain the possible applied optima, we would not be talking about a derivation of applied best practices from the eternal design. Instead, the transition would be purely conventional and arbitrary. So how can just human laws be derived from, but not entailed by, eternal law? The solution is fairly obvious. Eternal law does not entail just human laws autonomously, but only under an adequately specified set of boundary conditions. This happens to be the solution Aquinas adopts in his Treatise on Law (see qq. 95–97), but on reflection one also sees it is the only solution possible. There’s only one conclusion to draw: contrary to the popular impression, Aquinas does not understand eternal laws as rules of human conduct, but as high-level constraints on practical reasoning. They are normative operators that generate rules of conduct only by taking contingent inputs. The resulting norms are constructed and relative because their content depends on contingent circumstances, but they are also objective and absolute because their validity—under the specified type of boundary conditions—is eternal. Aquinas is the stand-by for almost all casual references to “natural law” in contemporary literature. However, a careful reading of the very questions usually cited (Summa Theologiae, Ia IIae, qq. 90–97) does not support the popular interpretation. Instead of extreme realism and inflexible absolutism, what we find is a much more complex position that manages to check all the meta-ethical boxes in Wong’s list of alternatives. ARISTOTLE. All the evidence I present about pre-modern natural law theory runs counter to its popular interpretation, but the next piece of evidence is a particularly potent bombshell. Although not usually Besides the well-known treatment in his “Treatise on Law” (op. cit.), Aquinas also discusses the relation of natural law to unchanging eternal law on the one side and changing human law on the other in Summa Theologiae, IIa IIae, q. 47, aa.10–12, and in De regno, ad regem Cypri (a.k.a. De regimine principum), Bk. I, chs. 13–16. 10 17 noticed, the same coincidence of opposites we found in Aquinas can be found in Aristotle. To see it, we need only to compare his stipulations about natural justice11 and remedial equity:12 1. “Just by nature” is that which is normatively valid everywhere. Written laws differ here and there; the naturally just is universal and invariable. 2. Because human situations are inherently variable and particular, there is no such thing as a rule that applies everywhere. Justice will often require remedial equity, that is, adjustment of the rule in light of particular circumstances. The tension between (1) and (2) is patent. If there is such a thing as natural justice, it should not need any correction by equity. Conversely, if there is no such thing as a universally valid rule, then nothing can be just by nature. Yet (1) and (2) are both non-negotiable commitments for Aristotle—and for the rest of the natural law tradition. No one sees any contradiction between them. An easy solution might look tempting: natural justice is generic, so equity must see to the specifics of application. But this solution entails that equity itself be purely conventional, when everyone knows that is exactly what equity is not. Considerations of fairness have always been the paradigm example of jurisprudence that is supposedly based on “nature” and “reason” rather than convention. When the judge bends the rule to accommodate mitigating circumstances, she is doing it in the conviction that any judge—even from another place and time— would, in her situation, rule exactly the same way. Somehow the universality is in the specificity, not opposed to it. Those familiar with Aristotle’s ethics will recognize this particular conceit. It’s his doctrine of virtuous action as a mean between the extremes of excess and deficiency. Since what counts as excess or deficiency depends on the specifics of the situation, what counts as the mean does, too. It follows, since no two situations are exactly alike, that the right thing to do is relative—it’s relative to the uniqueness of each and every situation. Nevertheless, action can be rationally decided on objective grounds because there is, in each situation, exactly one right thing to do. Rightness is not relative in the subjectivist sense that mutually incompatible judgments of rightness can each be true.13 So even Aristotle answers “all of the above” to Wong’s cardinal meta-ethical questions. Virtuous behavior constructs a response that is relative to a unique situation, but it seeks to do the one thing that is inherently right in that situation. The judge’s job is the complicated one of constructing what the just-right response to a situation would be without the benefit of experiencing it first-hand. While her information is more limited, her basic competence is the same as that of the agent actually involved. The doctrine of the mean is Aristotle’s generic solution to the paradox we found at the heart of practical reasoning (pp. 6–12 above). It allows him to grant the adventitious character of experience and nevertheless affirm that conduct can be justified sub specie aeternitatis. FRANCISCO SUAREZ. Judgment “tailored to the circumstances, but not subjective” is the familiar mission of jurisprudence. I’m suggesting it’s the key to all natural law theory and, in its jurisprudential form, the key to all absolutist approaches in natural law theory. 11 to dikaion physikon: Eth. Nic., 1134b18–1135a5. 12 epieikeia: ibid., 1137a31–1138a3. 13 See Eth. Nic., Bks. II, III, & VI. 18 To be clear, judgment “tailored to the circumstances, but not subjective” is equally key on my interpretation to the utilitarian approach. The difference in this case is that it doesn’t need to take a jurisprudential form. Hobbes gave us a taste of this, and later exhibits will make it abundantly clear. Since utilitarians model their natural laws on successful human practices, their laws are tailored to real circumstances from the outset. Consequently, although they often ramify their social theories with any number of applications, there’s no place for jurisprudential thinking in the initial construction of their theories. As we will see, it’s meeting the “not subjective” condition that animates the utilitarians’ distinctive kind of theorizing. But it’s the alternative approach that concerns us now. The pre-modern tradition of natural law perfectly illustrates the dilemma absolutist approaches always face. The problem is that principles are meaningless until the realities of their application are worked out. In other words, the jurisprudence of application was always the beating heart of realism and absolutism in natural law theory. Perhaps this sounds like an overbold claim, but I find compelling evidence in Francisco Suarez, the last and most exacting of Scholastic philosophers. He devotes a whole questio in his Tractatus de legibus to the relation between Aristotle’s remedial equity (epieikeia) and natural law.14 Suarez denies that primary natural laws could ever need correction through remedial equity—if it looks that way, it’s only because you have misconstrued what the law in question applies to. (A law is universal with respect to what does—not what doesn’t—fall within its scope.) At the same time, however, Suarez insists that primary natural laws are highly abstract. It’s never immediately evident how and when abstract laws apply to individual circumstances. Consequently, a process of complex reasoning must always intervene: Therefore, in order to understand the true sense of a natural precept, we must inquire into the conditions and circumstances under which the act concerned is essentially good or evil; and this inquiry is spoken of as the interpretation of a natural precept with respect to the true sense of the precept.15 Suarez understands the process that determines correct application to be entirely rational. At the same time, it’s an accommodation that has to take the unique specificity of the empirical situation into consideration. This is the same transmutation we have seen before. Starting with the contingent event and describing its singularity, we end up with the unique type of that event-token. I’m making it sound like alchemy, but the procedure in question is ordinary jurisprudential reasoning. It determines with certainty whether a given real action is permitted, forbidden, or obligatory on the grounds that such an action would always be of such a character under the conditions specified (“it is never wrong/never right to do/forgo an act of such and such a type under such and such conditions”). Nevertheless, there is something peculiar about jurisprudential reasoning in general, and Suarez’s nonstandard language bears witness to it: “we must inquire into the conditions and circumstances (conditiones et circumstantias) under which the act concerned is essentially (secundum se) good or evil.” Properly speaking, “essential” is what doesn’t depend on “conditions and circumstances.” So, unless Suarez is contradicting himself, he must agree with Aristotle that the essence–accident distinction breaks down when it comes to the beings called “actions.”16 Actions don’t have essences. They are one14 Suarez, Tractatus de Legibus, Bk. II, q. 16. 15 Op. cit., II.16.6. Translation from: Selections from Three Works, ed. Thomas Pink (Carmel: Liberty Fund), 2015, pp. 360–1. 16 The doctrine of the mean is predicated on the singularity of all actions; see Eth. Nic., Bk. II. 19 off events. So what they are “according to themselves” (secundum se) is nothing less than the unique totality of what they are in each concrete context. What Suarez is therefore saying in the quoted passage is that we know the “true sense” of eternal, unchanging laws only when they are fully interpreted in light of changing, contingent circumstances. The conclusion Suarez reaches is simply a converse statement of the principle we found in Aristotle. Aristotle said the right thing for a particular person to do in a particular situation is the one thing any such person in exactly that situation should always do. Suarez says that what one should always do differs depending on the exact specifics of the total situation. In each case we see the same coincidence of opposites. To be unconditionally valid, a natural law cannot be abstract. It must be articulated in terms of a specific type of real situation. Keep in mind what that means. Unless we fudge, each type will be so specific as to have only one actual token. Fortunately for human beings, many situations are similar enough to permit fudging. But strictly speaking, it is only in relation to a singular event that each articulation of natural law is “eternally” valid. What must we say about that kind of natural law? Is it absolute? Yes. Is it relative? Yes. Is it objectively real? Yes. Is it constructed? Yes. Another relevant question would be: is it coherent? For the moment I note only that if we decide it isn’t, we would have to draw the same conclusion about Aristotle’s theory of virtuous action and probably, too, about Aquinas’ theory of just human law. WILLIAM OF OCKHAM. Based on the writings of William of Ockham, a strong case can be made that natural law has always been understood the paradoxical way Suarez understands it. The same coincidence of opposites we find in Suarez is evident in Ockham’s idea of “conditional natural law”—a law that is eternal and contingent at the same time. But this idea has a long history. It appears already in Augustine, whose authority guaranteed its influence throughout the Middle Ages. It reappears in modern natural law theory as the “rational construction” of changing applications and equitable exceptions. And it lives on today in any jurisprudence that recognizes a difference between the letter and the spirit of the law. The long continuity I’m alleging may give readers pause. But that’s why I propose now to look at William of Ockham. What if Ockham, for all his notoriety, was not the outlier he is usually reckoned to be? Won’t the rest of the natural law tradition start to look different, too? Let’s begin with the main point of controversy: Was Ockham a die-hard opponent of Scholastic orthodoxy? Many scholars think so. They are convinced that the extreme theological voluntarism for which Ockham is famous reduces any kind of law—a law of physics no less than a law of morality—to an arbitrary command of God.17 It seems to follow from his theology that laws should never be understood as absolute and real, but as constructed and relative at least to the acts of will that create them. Of course, this construction is referred to the singular divine will, so we are still a far cry from an “anything goes” subjectivism. But we also seem to be quite far from the paradigmatic realism and absolutism of Aquinas, for whom the source of law is the divine intellect. Intellect operates by discovering a truth. A discovered truth is one that already exists, independent of the mind. In the special case that interests us, intellect discovers the truth about what’s good and bad given the complexities of a possible creation. For Aquinas, even God’s will must acquiesce to the primordial truths His intellect discerns in contemplating the possibilities of creation. For Ockham, it’s the other way around. It’s a fundamental article of faith that God is omnipotent. So good and bad can’t be things that constrain God. He creates 17 Best known for promoting this line of interpretation in English-language scholarship is Francis Oakley, in French Michel Villey, and in German Hans Blumenberg. 20 them. It’s intellect that must acquiesce to whatever God happens to will. On this telling, Ockham’s voluntarism is a stealth weapon intended to destroy any vestige of realism or absolutism in Scholastic philosophy. But things are not that simple. It is very well known that Ockham, despite illustrating his theological arguments with rather extreme examples (e.g., the possibility of God commanding creatures not to love Him), nowhere draws the skeptical conclusions for applied natural knowledge that modern readers, based on his voluntarism in theology, expect him to draw.18 However puzzling it may be that Ockham’s voluntarism did not lead him to adopt an extreme skeptical position on morals, we know for a fact that it did not.19 His voluminous writings on political philosophy, which occupied the last two decades of his life, reveal, despite their radical take on ecclesiology, an unquestioned baseline of conservative continuity with traditional natural law theory.20 Moreover, such conservativism is not limited to his later writings. At the same time he was provocatively insisting on the ascendency of God’s will over His intellect, he was also maintaining the traditional Scholastic position that moral truths were not contingent, but “scientifically demonstrable.”21 However, if Ockham was committed to an in-principle realism and absolutism about natural law—both in his Oxford period and in his political writings—then the equally evident strains of relativism and constructivism are all the more curious. What to make of this countervailing tendency? We miss its true significance if we put it down to Ockham being a renegade who challenged Scholastic tradition with revolutionary ideas. He did no such thing. He was a renegade in papal politics, but—despite his innovations in theology—not in moral philosophy or law. So where do Ockham’s sympathies with relativism and constructivism come from? They come from nothing other than his baseline commitment to absolutism and realism. We saw the same thing with Aquinas and Suarez. Commitment to absolutism and realism does not exclude concessions to relativism and constructivism, but—under the rubric of “application”—actually entails them. Ockham is not an outlier in this regard. Relativism and constructivism are typical features of mainstream Scholasticism, predicated on—not against—an underlying commitment to absolutism and realism. Even if I’ve managed to convince a skeptical reader that my interpretation of Ockham and his place in the history of natural law theory is coherent, I still haven’t offered any evidence for my unorthodox view. So what’s the proof? It’s sufficient for our purposes to point to Ockham’s notion of a “conditional” (ex suppositione) natural law that was changeable22 and to notice that Ockham was not innovating, but Ockham’s baseline orthodoxy (his adherence, that is, to some form of Aristotelian rationalism in morality) is documented by Peter King (“Ockham’s Ethical Theory”) and Marilyn Adams (“Ockham on Will, Nature, and Morality”) in their contributions to The Cambridge Companion to Ockham (Cambridge: University Press), 1999, pp. 227–244 and 245–272, respectively. 18 The untenability of interpretations in the Oakely–Villey mold is well demonstrated by Brian Tierney, The Idea of Natural Rights (Atlanta: Scholars Press), 1997. See especially chs. 1, 7, & 8. 19 Skeptics should see Steven Barbone, “Natural Law in William of Ockham,” International Studies in Philosophy, XXVIII.2 (1996), pp. 19-34, and A.S. McGrade, “Natural Law and Moral Omnipotence,” in The Cambridge Companion to Ockham (op. cit.), pp. 273–302. 20 21 That Ockham did not develop orthodox views only after his dramatic flight from Avignon and sudden intellectual turn, as outlaw enemy of the Pope, to exclusive engagement with political theory is conclusively shown by his Oxford Quodlibets. Quodlibet II.14, for example, is dispositive. Ockham matter-of-factly concludes that a demonstrative science of morals is possible. 22 Ockham, Dialogus, Pt. III, Tract II, Bk. 3, ch. 6. 21 drawing on a tradition extending at least as far back as Augustine.23 For Ockham, there are, to be sure, immutable natural laws (he gives the commandments against adultery and lying as examples). That nonnegotiable conviction establishes his bona fides as an absolutist and realist. However, some natural laws—such as property being held in common and all humans being free and equal—were valid only in the primordial state of innocence. Once human beings committed sin, the demands of equity necessarily changed. If there is assault, for example, repelling force with force becomes equitable. If there is uncharitable hoarding of common goods, private property becomes equitable. Most important, even the “immutable” natural laws are not wholly immune to change since what counts as an instance of adultery or lying will depend, at least in part, on contingent circumstances and on positive laws prevailing at the time. It’s hardly revolutionary. Ockham’s natural law theory is clearly modeled on Aristotle’s doctrine of the mean, which makes all virtuous action situationally unique. The case is compelling: what’s “naturally just” couldn’t possibly be inequitable. So how could natural justice not be something that—in each and every case—takes the totality of actual circumstances into account? The meaning of natural law is always determined on condition (ex suppositione) of the particular circumstances in which it is to be applied. The result, however, is paradoxical: natural justice is the “same everywhere” only by changing from place to place. By the same token, jurisprudence becomes indispensable to the epistemology of natural law. No one could possibly know what’s naturally just in any instance without actually going through the process of taking all the circumstantial particulars into account.24 The absoluteness of God’s commands (e.g., the Decalogue) means they must be the starting point in the determination of what’s right, but the particularity of justice means they can’t be anything more than just the starting point. If that sounds like an unorthodox position for a Catholic theologian to take, we need to recall that it is the same position taken by Jesus, as the story of his healing a sick man on the Sabbath pointedly illustrates (Mark 3). Ockham’s idea of a conditional natural law that is changeable—that meets the same eternal standards by adapting in appropriately different ways to different circumstances—does not just reappear in Suarez (the conservative tail end of Scholasticism). It passes smoothly into the modern tradition as “adventitious” natural law (Pufendorf), or as law that is natural not “properly,” but “by reduction” (Grotius, Pufendorf), or simply as “secondary” natural law (Burlamaqui).25 These are all different names for what I described above, when discussing Aquinas, as eternal law applied to contingent human circumstances. And it cannot be stressed enough that this idea was always mainstream—it’s an uncontroversial part of Catholic natural law doctrine in the Middle Ages. It already figured prominently in Augustine’s discussion of eternal law in Part One of The Free Choice of the Will26—a standard reference point for all Scholastic legal theory—and reappears, prior to Ockham, in canonist commentaries on Gratian’s Decretum.27 23 Augustine, De libero arbitrio, I.vi.50. Op. cit., Pt. III, Tract II, Bk. 1, ch. 15. The impossibility of ever being sure you have taken all the relevant circumstances into account introduces an element of fallibility into jurisprudence that Ockham is implicitly recognizing in this chapter. Such fallibility of jurisprudence was already stressed to dramatic effect by Augustine (see De civitate Dei, XIX.6). 24 Grotius, The Rights of War and Peace, I.i.x; Pufendorf, The Law of Nature and of Nations, I.i.7; II.iii.22; III.iv.3; Burlamaqui, The Principles of Natural Law, II.4.xxiv. 25 26 Loc. cit. 27 Tierney, op. cit., pp. 176–182. 22 The thrust of these exhibits from Aristotle and Scholasticism is that “eternal law” (which supposedly has “the same” validity everywhere) has no definite prescriptive implications for human behavior until it is applied in contingent circumstances. An unapplied eternal law is perhaps not even a rule of action. It seems to express an axiological priority that regulates jurisprudential reasoning about action rather than action itself. The New Testament passage already cited offers a perfect illustration. It’s coincidence that Jesus encounters a man in need of charity on the Sabbath, but that coincidence forces us to revise our understanding of the Third Commandment (to observe the Sabbath and keep it holy). Either we treat its validity as conditional on not conflicting with a superior requirement of charity, or we add “practicing charity” to what keeping the Sabbath holy means. Either way, it’s only by reflecting on the implications of its presumptive application in different circumstances that can we work back to the true scope and meaning of the Third Commandment. A good analogy for “applied eternal law” might be melted snow. It’s true that melted snow both is snow (it’s snow that has melted) and is not snow (it’s water). In much the same way, natural law both is eternal (it’s eternal law that’s been applied) and is not eternal (once applied, it’s variable). Insofar as it’s derived from eternal law, it’s absolute and objective. Insofar as it’s applied to changing circumstances, it’s relative and constructed. The crucial thing is that “application” does not mean absoluteness gets left behind. As long as it’s the one appropriate derivation under exactly these circumstances, the application retains the absoluteness of its source. I’ve tried to capture that in my analogy. The melt-water is different from snow because it’s fluid and accommodates itself to the shape of the basin that receives it, but it isn’t a different substance from snow. It’s the snow itself at a nonfreezing temperature. That’s why I speak of a coincidence of opposites. We see this clearly in the recurring idea that an applied natural law articulates what would, in a given (and possibly unique) historical situation, “always” be the right thing to do. The eternally unchanging character of natural law is still present in each of its historically changing articulations. The traditional job of jurisprudence is to certify the peculiar coincidence of “valid only once” and “valid forever” that defines “application” by providing a persuasive account of how we can derive the law in question from our abstract and supposedly unchanging principles simply by taking circumstantially relevant constraints into consideration. Jack, who featured in my parable about practical reason, is engaged in something quite similar to traditional jurisprudence when the experience of calamity teaches him to be more prudent in the future. From a possibly unique experience, Jack also learns something that is “valid forever.” It’s eternally valid in the sense that it will be an appropriate response for every possible recurrence of the same circumstances. The fact that the exact same circumstances will not recur is irrelevant. Something similar will usually be appropriate in similar circumstances. Of course, the exceptions to what’s usual cannot be foreseen. So, in the face of unexpected calamities, the maxims of practical reason will evolve. Jack will revise his rules for success the same way jurisprudence continually reinterprets law in light of novel circumstances. Far from being an outlier in the long tradition of natural law, Ockham gives us an account of practical reason that is paradigmatic. The process it describes explains everyday reasoning as well as formal jurisprudence. It accurately characterizes the jurisprudential element in both modern and pre-modern theory. Most important, it explains why absolutist approaches always have recourse to jurisprudence and why recourse to jurisprudence always makes natural law theory “all of the above.” 23 EDMUND BURKE. Burke is famous for his virulent denunciations of the Enlightenment’s preoccupation with natural law and natural rights. Consequently, despite his many insistent references to what’s “natural” in matters of policy, he is often considered natural law’s most unforgiving enemy. But that can’t be right. Explicit appeals to “natural laws” and “natural rights” as ultimate normative principles, while not legion in Burke’s writings, are nevertheless impossible to miss.28 In the Enlightenment’s rejection of tradition, authority, and custom, Burke saw a steep slippery slope that he thought could lead only to nihilism. Far from grounding us, rational critique deracinates us from the only ground there is: that certified organic product of social evolution we know as our inherited culture. Burke is sure: it’s the examined life that will prove not worth living. So it goes without saying that Burke opposed the rationalistic theories of natural law that were considered state of the art in the 18th century, and of course he opposed the French Revolution with its literal formulary of abstract rights based on reason. However, he based his opposition to these things on what he purveyed as a correct understanding of natural law. Burke didn’t deny that we can rationally identify “rights of man.” On the contrary, they are entailed by the purpose for which society was founded (RRF 217–8).29 But he insisted on two points. First, in agreement with the position we found in Ockham and Suarez, he insists that natural laws have meaning only insofar as they are applied in real circumstances (RRF 151 & 220). Second, in opposition—as far as I can tell—to the entire natural law tradition, he insists that reason is not competent to carry out such an application. Society is an artifact, to be sure, but not an artifact of rational design. Accordingly, its “right” functioning is not the purview of reason. Reason cannot even fathom—never mind replace or improve upon—the ineffable wisdom embedded in long-tried and incrementally-evolved social practices. Just like a leather shoe you have worn for a long time, the mores of each society fit it perfectly. That perfect fit is something one experiences. It cannot be rationally reconstructed or theoretically explained. So there is such a thing as “applied natural law” for Burke, but it’s the anonymous forces of social history that apply it. It follows that the only way we know it is a posteriori—as something incumbent in the traditional and customary bases of sustained social order (RRF 184–5 & 219–220). This is the source of Burkes’ conservatism. Tradition is sublime and irrefragable. Deferring to it is intrinsically right. Justice is preserving the patrimony. Clearly, norms for Burke are constructed by human beings and defined relative to their historical situations. But Burke also thinks they are objectively real. How could they not be if norms are rules that reflect precisely those patterns of adaptation that have been successful? So Burke is convinced that, given the real circumstances a particular community has evolved in, existing norms are objective facts about what will and will not succeed at promoting social well-being. Like it or not, his argument is not without substance. He need only stress (as he always does) how dependent “promoting social well-being” is on maintaining social order. Of course, the key to maintaining social order is for the most part a matter of maintaining the existing social order. That’s always Burke’s trump card: the existing order is part of the real environment in which we seek to promote well-being. But that’s not his only card. On his understanding of social evolution, it’s never wise to try second-guessing the utility of existing conventions. That would be like thinking new shoes will fit Those skeptical of my claim should see Peter Stanlis’ 1958 monograph, Edmund Burke and the Natural Law (Ann Arbor: The University of Michigan Press). 28 29 References are to Burke, Reflections on the Revolution in France, ed. Clarke (Stanford: Stanford University Press), 2001. 24 the unique shape of your feet better than your old ones. So, despite being historically contingent, conventions embody norms that are, without question, objectively real. The implication is clear. Burke’s answer to the question, “Are norms constructed or real?” is an unapologetic both. Can we go further and say that for Burke norms are also absolute? Indeed. For Burke there is nothing subjective about it. The norms of each society are absolutely right for that society. Given Burke’s antirationalist assumptions, it goes without saying that this is not something he could prove in any particular case, but it does follow generically for all cases, given his assumptions. Most notable for my purposes: he bolsters this view with an explicit reference to the Aristotelian doctrine that the right thing to do is different for every situation (and so relative to it), but absolutely right for each (RRF 221). Consequently, even to the question, “Are norms relative or absolute?” Burke’s answer is an unapologetic both. On the one hand, norms are valid only for the society in which they evolved. On the other, the evolved norms of any society are absolutely right for it, regardless of what anyone thinks about the matter. In fact, this statement of particularism is itself a universally true principle for Burke. Since traditions differ, the values conservatives hold will necessarily be different in different societies. Nevertheless, conservatism is always right, radicalism always wrong. RESULTS OF THE FIRST SERIES OF EXHIBITS. This concludes my first series of exhibits. How homogenous is the group we’ve surveyed? Let’s begin with the obvious. If we bracket Hobbes and Burke, there can be no question that all the figures considered so far (Aristotle, Aquinas, Ockham, and Suarez) endorsed a remarkably similar concept of natural law. Natural law is a context-sensitive rule for optimizing social well-being. It’s based partly on universalizing theory (roughly: what’s invoked as “eternal”) and partly on experience (whatever conditions the “application”). Notice that neither of these two bases can function without the other. So it’s no accident that all our pre-modern authors characterize natural law in a way that suggest an “all of the above” answer to Wong’s meta-ethical questions. Their concept of natural law is predicated on a dualism of opposites. Only a theory that was “all of the above” could possibly navigate between those two extremes. However, natural law theory does not end with Spanish Scholasticism. At the same moment Suarez was crowning the long Scholastic tradition, natural law was being reborn in a new form to a spectacular new career. As representatives of modern natural law theory, I included Hobbes and Burke among my exhibits. Burke’s anti-rationalism would seem to put him at odds with the Scholastic tradition. There’s no denying that in many ways it does, but the contrast is not as stark as popular caricatures suggest. We have seen that Scholastics never took the simple, one-sided positions so often attributed to them. In seeking a normative theory that combines and reconciles “absolute” and “relative,” Burke continues a pattern we found deeply imprinted throughout the tradition of Scholasticism. Most striking is the fact that Burke, just like the Scholastics before him, embraces Aristotle’s doctrine of the mean as the way to achieve that reconciliation. For Burke, of course, it’s not a rational reconciliation achieved by the jurist, but something always already accomplished by the practical pressures of social history. If we are looking for an odd man out, it has to be Hobbes—but not because he doesn’t seek the same reconciliation as the others. It’s just that he seeks it in a totally different way. Hobbes eschews any recourse to the Aristotelian idea that action is always situationally unique. For all the others, natural laws must be constructed (whether by reason or by history) because they need to be continually tailored and retailored to fit novel situations. Hobbes, by contrast, is quite certain that one solution fits all possible cases. The challenge for Hobbes is that nature is a realm of violence. It offers us no hint of a 25 clue—never mind a blueprint—for how we might live together in peace. Reason alone can solve the problem. It must create from scratch what nature fails to prefigure. But since the challenge nature presents admits of only one possible solution, reason is actually discovering something objectively real. I included Hobbes and Burke in my survey because, while ensuring the diversity of my sample, they deliver the strongest possible evidence in favor of extending the analysis I gave of Aristotle and the Scholastics to modern natural law theory. There’s no denying that Hobbes sees norms as both constructed and objectively real, no denying that Burke sees them as both absolute and relative. We misunderstand these meta-ethical tensions if we think they are peculiar to thinkers willfully breaking with the orthodoxy of the pre-modern natural law tradition. On the contrary, they are pronounced manifestations of something incipient in natural law theory from the first. The question we must now address is what that something is. What deep commitment does the modern phase of the natural law tradition still have in common with the pre-modern phase that dictates the same “all of the above” outcome? Despite many differences that do separate modern from the pre-modern natural law theorists—not least of which are the extreme outlying positions of Hobbes and Burke on epistemology—we still find an astonishing amount of baseline agreement. All the figures we have considered take it for granted that norms satisfy an urgent societal need—the need for order. Allowing for significant differences in their concepts of order, we can generalize broadly: for everyone we’ve considered, saying a norm is authorized by “natural law” just means it’s successful—not once off, but more or less reliably—in meeting that need. Of course they have different concepts of success, different ways of measuring it, and different ways of explaining how it comes about and how we can know what it is. But for each and every one of them, actual success in facilitating social well-being is the hard core of what it means for a norm to be “natural.” In fact, that’s precisely how we know every figure on our list truly is a natural law theorist. Despite significant elements of positivism in the legal thinking of both Hobbes and Burke, the fact that success alone can validate certain norms (and that failure can sometimes invalidate positive ones) proves that both of them are, at bottom, natural law theorists. This much should be relatively uncontroversial. But to find what I take to be the most important commonality across the tradition, we must venture into less certain territory. What I have tried to make evident in my exhibits is that everyone in the survey, regardless how they are typically identified or actually self-identify, in fact believes that norms are constructed. Construction is not to be confused with application, however. It’s true the exhibits show again and again how the need to apply general rules to specific situations creates an epistemological problem that forces realists and absolutists to compromise with contingency. But that’s not necessarily the case with those natural law theorists who take a utilitarian approach. Certainly the problem of application reappears in modern natural law theory. We see it in Grotius and Pufendorf, for example. Indeed, we see it in any theory that employs jurisprudence. But application plays no role in the natural law theory of Hobbes, and it plays an entirely different role in Burke’s theory, where it’s a fait accompli of social evolution, not a challenge for jurisprudence. Nevertheless, for Hobbes and Burke both, natural laws are most certainly constructed. In one sense, I’m making an obvious and trivial claim. Hobbes and Burke both reject the Aristotelian idea that polity exists “by nature,” so of course they think norms are constructed. However, the important thing I demonstrated in the exhibits is that neither of them see rejecting natural teleology as a reason to reject universalizing social theory or the idea of unconditional principles. To say, with Hobbes, that the 26 irresistibility of sovereign power is desirable always and everywhere is to say that it is an absolute good. Tradition plays the same role for Burke, entailing the absolute necessity of such human institutions as ownership of real property and inheritance of rights. It’s not by poor choice of words that Hobbes and Burke continue to talk about natural law. On the contrary, they retain a mostly classical view of what it is. So the epistemological problem of applying the general to the particular is one reason why natural law theorists think natural laws must be constructed, but it’s not the only reason. Those who reject natural teleology must reject the eternal verities it supposedly underwrites. But they don’t reject the idea of ultimate normative truths. Instead, they look for ways to generate them—to construct them, under some appropriate set of constraints, out of contingent truths that, as it happens, no one disputes. For Hobbes, normative truths are rational constructs. For Burke, they are empirically known constructs of social evolution. But, one way or another, some of them end up being as universal and absolute as any natural law of the pre-modern tradition. It’s not even clear what grounds traditional absolutists would have for disputing my assessment once we see that their natural laws—because they, too, required construction—were never as universal or as absolute as conventional interpretation makes them out to be. In conclusion, then, the crucial thing the exhibits have shown is that natural law theorists have different rationales for why natural laws are constructed, but—without disagreeing that much about what natural laws are—they do all agree that natural laws are constructed. To our contemporary, meta-ethically enlightened way of thinking, that would mean norms are invented, and their validity would be relative to those who felt the need to invent them. But for none of our authors do those conclusions follow. As the exhibits show, despite differing in their rationales, they all see the construction of norms as the manner in which objectively real and absolutely valid solutions are discovered. Whether a position so paradoxical by our lights is ultimately coherent is a question I defer to the conclusion of this paper. What’s certain is that western natural law theorists of all periods and persuasions incline to an “all of the above” type of meta-ethics. The persistence of that view may be the result of a confusion, but it’s not the result of an accident. They all appear to be motivated by the same more or less conscious conviction: construction is not the enemy of realism. What results from the contingently-motivated “artifice” of human endeavor is not necessarily “factitious.” Under appropriate constraints, construction is the vehicle of objectively real cognition. If what I said before commencing the exhibits was on point (see “A Parable about Practical Reason,” “Philosophical Implications of the Parable,” and “Forestalling Some Obvious Objections,” pp. 6–13 above), then what I’m describing is not a conceit of high learning or of antiquated world views we’ve now outgrown. It’s a conviction natural law theorists share with today’s man in the street. It shows up in the idea—familiar to anyone who’s learned caution from calamity—that experience is a coincidence of opposites, adventitiously yielding insight that is usually valid for all cases of similar type. But what must we now say? Does this filiation between natural law and common sense vindicate the former—or damn the latter? Does it prove that natural law is coherent because common sense is, or prove that common sense is not coherent because natural law isn’t either? I’m not sure. I’ll venture a tentative answer in my conclusion. But we must accept that it is one or the other. CONCLUDING EXHIBITS: PLATO AND PROTAGORAS. As capstone to my catalog of exhibits, I want to examine Plato and Protagoras in tandem. Although the one was an avowed absolutist and realist about norms, the other an avowed relativist and constructivist, there’s no question that they were both fully- 27 committed natural law theorists: each of them achieves rigor in practical philosophy by prosecuting a maximally formal analysis of the objective conditions of human functioning. If Plato and Protagoras are meta-ethical antipodes, their common employment of the natural law paradigm would demonstrate its versatility. However, the weight of the evidence presented so far tends in a different direction. What their shared commitment puts on display is not the versatility of natural law, but its complexity, and what it proves is that Plato and Protagoras are not meta-ethical antipodes. If my interpretation is right, Plato should prove to be as much (or almost as much) a relativist and constructivist as Protagoras, and Protagoras as much (or almost as much) an absolutist and realist as Plato. THE YOUNGER PLATO. In the Republic, Plato looks like a classic natural law theorist. He employs arm-chair reasoning to derive the constitution of his ideal state from the nature of the human soul and the conditions of its optimal functioning. Beginning with a rough but good-enough conceptual analysis of the structure of society, Plato infers the cause of that structure, which is the unchanging structure of the human soul. Knowing how the soul is structured then enables Plato to infer what kind of organization would make society function best (it’s the one most in conformity with the structure of the soul) and to explain why real societies, which are variously organized, have the careers they do—some succeeding, others not, some failing in one way, some in others. He can show not only how things that actually came to pass were, under the existing forms of polity, necessary outcomes, but also how they could have been avoided under a different form of polity. Plato thus appears in the Republic to be in possession of perfect political prudence. At the foundation of everything in the Republic is the tri-partite nature of the soul. Human behavior is driven by three kinds of motives: carnal appetite, social emotions, and reason. Optimal functioning requires the soul to be at peace with itself, and that requires appetite and emotion to obey reason— one’s own if that is possible, someone else’s if it is not. The purpose of the state is to put those who do not obey their own reason under the direct authority of those who do. The lesson of the Republic is that the conditions under which the individual is at peace with herself and the conditions under which society is at peace with itself are exactly the same set of conditions. THE OLDER PLATO. In a later and even longer dialog, Plato takes a different approach, both to social theory and to moral psychology. In the opening books of the Laws, Plato sounds more like Jeremy Bentham than the author of the Republic. A human being, he says, is like a puppet actuated by two strings—fear of pain and desire for pleasure (644c–645b).30 A moderating influence is possible, however, because our reactions are not limited to present pleasure and pain. We can calculate in advance the aggregate desirability of different outcomes. Right action is conduct that leads to the calculated best outcome. A law, we are told, is what we call it when a whole community agrees on one of these calculations. Notice that Plato has replaced the harmonization of faculties, which was the normative foundation of the Republic, with popular consensus about the long-term maximization of pleasure. In the fourth book, Plato has the Athenian Stranger announce categorically: “no human being ever legislates anything … chance and accidents … legislate everything for us.”31 He cites war, poverty, and epidemics as the true legislators. He means to stress thereby that laws are constructed in response to See Bentham, Principles of Morals and Legislation: “Nature has placed mankind under the governance of two sovereign masters, pain and pleasure” (I.i). 30 31 De Legibus, 709a–b. See The Laws of Plato, tr. Thomas Pangle (New York: Basic Books), 1980, p. 94. 28 adventitious events, and their value is relative to the exigencies we suffer because of those events. There is no talk of deriving laws or constitutions from the underlying constraints of human psychology. Has Plato rejected the realism and absolutism that was so vividly on display in the Republic? Not at all. The Stranger tells us: from experience comes skill (technē) in negotiating contingencies (709c). The possibility of skillfully negotiating them puts contingencies in an entirely new light. In one sense they are caused by chance (what we call “acts of God”), but in another sense they are caused by the negligence or ignorance of human beings. Plato’s realist intentions are evident when the Stranger describes the ability to avoid or mitigate misfortune as “being in possession of the truth” (709c: ἀληθείας ἐχόμενον). If Plato has not abandoned his earlier realism and absolutism, why the sudden preoccupation with adventitious experience? He repeatedly stresses how a good lawmaker, to calculate the best outcome, must learn from experience—both by studying examples from the past and by prudently learning the unpredictable lessons of the hard knocks that descend in the form of calamities and bad luck. Why the relentless insistence on the irrational provenance of practical wisdom? It’s because Plato knows that rational government of practical affairs involves a paradox. He already embraced that paradox when he said in the Republic that the Good was “beyond Being.” If the Good is beyond Being, then in order to be (to actually exist) it must become something other than good. In normative ethical theory—to take the obvious example—one is trying to give non-contingent justifications for contingently motivated behavior. However reasonable that may be, it’s still paradoxical. If we could really “eliminate” the contingency, there would be nothing in need of justification. We never close the gap. To be good, behavior must be non-contingently justified. To be, it must be contingent. There’s no question that Plato makes this point more emphatically in the Laws than he ever did before. He even has the Athenian Stranger tout the benefits of immoderate drinking. As rationale, he serves up a jarring eye-opener: A person may know what virtue is and even behave virtuously because she has been taught the truth about the nature of the human soul. But reading the instruction manual is a far cry from operational proficiency. To really know virtue, you have to experience what it takes to exercise good judgment and remain self-controlled even (or rather, especially) when disablingly intoxicated (645b–650b). The Stranger is adamant: in seeking rational control, there is no substitute for what we learn through the irrational test of actual adversity. It’s real life—not a hypothetical one we could construct without first experiencing it—that we must understand sub specie aeternitatis. In a way, the doctrinaire headiness of the Republic is a record of Plato’s own experience with intoxication. We can easily surmise it was the mindset leading to the disastrous political experiment in Syracuse. That experiment—bold in the making—must have seemed like utter folly after the fact. In the Laws, Plato speaks with the chastened wisdom of someone who grossly underestimated how far “beyond Being” the Good is and how much shabby Being you have to traverse before you can triangulate what the Good is “in itself.” The constructivism of the Laws is not a rejection of the realism and objectivism of the Republic, but of its naïve view that knowledge of the forms can somehow get ahead of knowledge of particulars. The same core insight seems to be at work in the reduction of “eternal law” to case-by-case jurisprudence that we found implicitly in Aristotle and Aquinas, explicitly in Ockham and Suarez. Might we find it in Xunzi, too? PROTAGORAS. The father of relativism used his theory of how all things are relative to establish criteria for justice that are absolute and universal. However contradictory that may be by the lights of our metaethical categories, Protagoras himself saw no contradiction in doing so. In a nutshell the argument is: 29 what’s “good” is relative to what any particular individual wants, and what’s “just” is relative to what any particular community is accustomed to, but no individual wanting to live in a community wants others to despoil her of life and property. That’s universal. Since people would flee any community that allowed such infringement, it’s also objective: a community not accustomed to honoring and enforcing mutual restraint would destroy itself “by nature.” Now one can object to the framing of this argument that if something is the same for everybody, then it’s not “relative.” But Protagoras need only point out that “same for everybody” is a relation and that it does precisely the same work as “objective and absolute.”32 In response, absolutists will object that there is no real necessity in the way Protagoras constructs his universal standard. What people happen to agree on is just another contingent and variable truth. But herein lies the genius of his argument. Even if the value individuals place on mutual restraint is contingent on their wanting to live in society, there’s nothing variable about it. People want to live in society for mutual benefit, not mutual harm. People don’t desire mutual restraint because it’s universally just. It’s universally just because it’s what everyone desires. In antiquity, it was especially the Epicureans who developed Protagoras’ political ideas. They were strident critics of the absolutism and objectivism purveyed by Plato and Aristotle. And yet, their relativism and constructivism did not lead them to skepticism or subjectivism about justice any more than it did Protagoras. In fact, the theory of justice Epicurus expounds in his Principle Doctrines (see §§31–38; Diogenes Laertius, Lives, 10.150–153) is not very different from what I extracted above from Ockham and Suarez. The convergence, in fact, is extraordinary: there is something “naturally just” (§31), it is common to all (§36), but it has no per se content33 since its specificity depends entirely on the unique contexts in which social compacts are made (§§33 & 37–38). Most important: given the social context, a positive law is either just or not just; opinion has nothing to do with it (§38). Constructed? Yes. Relative? Yes? Objective? Yes. Absolute? Yes. In conclusion, we have to say the same thing for Plato and the ancient contractarians that we said for Hobbes, Aquinas, Aristotle, Suarez, Ockham, and Burke: if we pose our two cardinal meta-ethical questions, the answer is all of the above. RESULTS OF THE CONCLUDING EXHIBITS. Protagoras and Epicurus never deny that the same solution will work again if the goals and the challenges are the same. No less than Plato, they seek norms that we can rely on—norms that are “always right” for reaching certain well-defined goals in certain well-defined kinds of situations. While that’s obvious from the sources, it’s not what conventional interpretations lead us to expect. But that’s only the beginning. Plato and the ancient contract theorists agree in even more surprising ways. Crucially, they also agree that what’s “always” or “eternally” right is a vacuous notion unless it’s constructed—under appropriate constraints—out of what people contingently find the need In the main, Protagoras’ views can be reconstructed from Plato’s Protagoras (320c–328d & 333d–334c) and Theaetetus (152a–172b). One must, of course, exercise due caution. Plato—at least at this point in his intellectual development—fears that any relativism leads to “anything goes” relativism (“whatever you believe is true for you”). He is therefore at pains to expose Protagoras as such a relativist. Plato’s invidious interpretation is easily dispatched. Protagoras did not say the gods exist for those who think they exist, and do not exist for those who think they do not. He famously said (Diogenes Laertius, Lives, 9.51) it was obscure, which means there is an objective fact of the matter that we don’t know. There is always an objective fact of the matter for Protagoras. If you say the wine was bitter, that’s true for you. If I say you found it sweet, that’s not “true for me”— it’s just wrong. How the wine tastes to you is relative to you, not to me. All things are relative, for sure, but they are not all relative to anything and everything. Each thing is relative to a determinable measure. 32 §33; Diogenes Laertius, Lives, 10.150: Οὐκ ἦν τι καθ᾽ ἑαυτὸ δικαιοσύνη. “There never was any such thing as justice in itself (or: as such).” 33 30 for. The relevant constraint for Protagoras is universal agreement on what we do and don’t want from society. For Plato it’s the irrational adversities to which we must adapt, both external (calamities) and internal (the kind of psychological realities exposed in drunkenness). In either case, they are aiming to construct a generally valid rule that will enable people to avoid the kind of calamities they experienced in the past. Protagoras presents as a relativist and constructivist in social theory because he sees the political response to life’s adversities as something ingeniously improvised. Plato presents as an absolutist and realist because he sees the response as something dictated by the human being’s unchanging need for security and the ultimate predictability of the challenges. It’s the same face-off we saw with Jack and Jill in my parable about practical reason. With this in mind, let’s take a quick glace back. One thing the first series of exhibits showed was that absolutists, sooner or later, concede the role adventitious experience plays in the formulation of their “absolutes.” They have to concede it because the only way to generate an actionable precept is to work out how the presupposed axiological priorities can be implemented in real world circumstances. The resulting application is patently contingent on the details of the circumstances in which normative guidance is sought. More than an “application,” it’s the construction of an empirically-conditioned rule. Although absoluteness is invariably claimed for it, the fallibility—indeed, the fragility—of the construction is evident in the way theorists readily revise a proposed rule as soon as critics show that following it would have negative consequences in situations the theorists failed to foresee. Of course, absolutists can concede all this and still claim to be legitimate absolutists on the grounds that their constructions are motivated by principles that are eternally and unconditionally right. But they cannot deny their curious habit of revising not just the applications, but even the principles, in light of problematic outcomes. In fact, both Ockham and Suarez are explicit that the “true” meaning of the supposedly eternal and unconditional principles cannot be determined apart from what we learn by trying to apply them. My point in stressing how absolutists actually proceed was to expose how different it is from how they advertise their projects or what the project of absolutism is reputed to be. In light of how they actually proceed in their theorizing, we are compelled to ask how much they actually differ from their supposed opponents, the utilitarians. Utilitarians embrace the idea that norms must be constructed. But that just means the utilitarian does on principle everything the absolutist does in fact. Fundamental differences between the two approaches became even more elusive when we realized that utilitarians, despite embracing the contingency of construction, aim at a social theory no less universalizing than absolutists. Hobbes famously rejects the Aristotelian theory of essences, but he underwrites his one-size-fits-all social theory with supposedly unchanging truths about human nature. Burke underwrites his particularism with the universal truth that tradition is the only source of value. In other words, utilitarians don’t just do on principle the same things absolutists do in fact, they also do in fact the same things absolutists do on principle. The difference between them seems to lie not in the kind of social theory they propose, but in what part of it they commit to “in principle” and what part they concede as an exigency in the course of execution. If the first series of exhibits raised the possibility that there are no truly fundamental differences between absolutist and utilitarian approaches, I think the concluding series clenches it. Plato is rare among absolutists because he doesn’t just concede back-handedly—in order to get on with the workaday business of “application”—that supposedly absolute norms are actually constructed. Even more than Ockham or Suarez, he insists on foregrounding the role of empirical construction in normative thinking. His point, in fact, is that empirical construction is an epistemological necessity for 31 normative thinking as such. But if that’s true, what’s left of the supposed differences between absolutism and utilitarianism? While they certainly differ in procedure and method, the principled opposition between them appears to be little more than rhetorical. I chose Plato and Protagoras for the concluding exhibits because they are not just paradigmatic of the two approaches. They are also paradigmatic of the vanishing difference between them. There’s no question that Plato and Protagoras employ different arguments and start from different principles and assumptions. But do they end up with theories that differ in their practical implications? As I noted above, Protagoras doesn’t think people desire mutual restraint because it’s universally just. It’s universally just because that’s what everyone desires. For Plato, of course, it’s the opposite. Justice is what agrees with the nature of society, which depends, in turn, on human nature. So people want it if they know what’s good for them. But once Plato begins to tell us how we know what the nature of society is, his differences with Protagoras begin to vanish. Plato asks, “What has contingent adversity taught us about the nature of society and the conditions of its success?” He means, of course, the “eternal” nature of society. That essentialist perspective is what qualifies him as an absolutist. However, by explicitly acknowledging the role of contingency, Plato has implicitly qualified his question: what has adversity taught us about the eternal conditions of success so far? That qualification makes all the difference. It’s hard to see how the question Plato is really asking differs from the utilitarian question Protagoras asks: “What kind of thing must we make society if, going forward, we want to avoid what displeased us in the past?” It’s clear that both philosophers hypothesize stable conditions of social success. The shared assumption is that past failures reveal the conditions a society will still have to meet in the future in order to be successful. Plato thinks that’s because society has an unchanging nature. Protagoras thinks it’s because societies are created by individuals who want the same thing (the benefits of close cooperation). But in what way do the two theorists actually differ? They both seek the success conditions of social life. Plato does not deny that the conditions we arrive at are hypothetical—that they are actually utilitarian generalizations that might need to be revised in the future. All he insists on is that there are stable conditions, that they are what we are aiming to formulate, and that we do—albeit gradually and imperfectly—learn about them by reflecting on the course of actual experience. Protagoras, on the other hand, does not deny that there are stable conditions. On the contrary, he predicates his theory on our ability to frame hypotheses—in the form of “this sort of thing or event always follows that sort of thing or event”—about what those stable conditions are (he calls this “wisdom”). Crucially, Protagoras doesn’t deny that some of our hypotheses may never need to be revised (his own theory being a case in point). But he insists that all we ever do as human beings is follow our desires. The only certainty about good and bad is that we are or are not satisfied by something. Whatever we learn about “this resulting from that” we learn on the fly in seeking to avoid what displeased us in the past. The remarkable thing is that Protagoras, in opposing absolutes, does not insist on anything Plato does not openly acknowledge in the Laws, while Plato, in seeking absolutes, does not insist on anything without qualifying it the same way Protagoras would. Practical Reason and Natural Law: Sketch of a Theory Natural law is a fancy name for practical reasoning. Everyday attempts to rationally criticize or justify behavior employ the same type of reasoning as the natural law philosophers. In everyday situations, the question is whether something you want or don’t want to do is actually the right thing to do. Doubt 32 manifests as anxiety about hypotheticals. “Is my keeping the money I found really ok to do? What if my mother finds out I did this? What if the person who dropped it really needs it?” The only difference between philosophy and everyday reasoning is the purchase power of doubt: when we start to include unlikely counterfactuals among our hypotheticals, we have become philosophers. The transition may be smoother than we think. To be confident, you need proof of rightness that answers doubt. That means having reasons for your choices that are unimpeachable. Unimpeachable is another way of saying “certainty there are no exceptions to what you say is the right thing to do.” In real life, of course, “unimpeachable” is a sliding scale. For some, it might mean “my mother wouldn’t think less of me.” But even the man in the street senses that what’s right for him to do in a particular situation would have to be right for anyone like him to do in such a situation. The philosopher is the one who realizes you cannot determine what would always be right for a particular kind of person to do in a particular kind of situation without recourse to norms that are not about particulars at all—that are valid for any kind of person in any kind of situation. The upshot of my analysis is that we are all natural law theorists, at least in a weak sense. Natural law theorists in the strong sense are the philosophers who seek unimpeachable confidence in practical reason through more or less explicit recourse to allegedly eternal and universal principles. To be clear: I don’t claim that can actually be done. But nor do I claim it can’t. I claim only that it’s the natural extension of what practical reason spontaneously aspires to. Notice that Burke, with his appeal to the universal sanctity of tradition, and Hobbes, with his appeal to the unchanging truth of human nature, are as much strong natural law theorists as Aquinas, with his appeal to eternal law. It might seem that the declared realists and absolutists are the ones who push the paradigm too far. But, as the exhibits show, they are the first to recognize that a transition from something eternally and universally right (e.g., happiness/peace is to be sought) to the right thing to do here-and-now (do/don’t hit him back) is as far as a thing can be from straightforward. We saw that a kind of casuistry (in the positive sense of the term) had to intervene and that natural laws were largely indeterminate until applied—ideally with the peculiar and irreplaceable expertise of those who know failure—to specific, real-world contexts. The result—which should no longer surprise us—is that even for the most conservative natural law theorists “objective and absolute” does not exclude, but actually entails “constructed and relative.” In light of these results, natural law looks less like the blunt instrument its critics decry and more like the shifting situational tactics they recommend. Perhaps in the final analysis there is no coherent method for making decisions about action based on “eternal and universal principles.” But that conclusion is far from obvious. The premise of “strong” natural law theory is prima facie preposterous only if we fail to acknowledge that even for absolutists and realists the principles are reinterpreted in light of each situation to which they are applied. I began this paper by characterizing two kinds of natural law theory: the typically pre-modern, according to which human society is something natural, and the typically modern, according to which human society is something artificial. We are now in a position to propose a better typology. The real difference in types of natural law theory is whether one puts the aspiration to universality up front, or brings it in later: 1. Some philosophers, persuaded that relativism and constructivism undermine the very possibility of norms, conceive of natural law more or less explicitly as eternal rules, but—apparently 33 without exception—to explain the knowability and relevance of natural law in real human time, they backtrack and introduce constraints that require natural law to operate with contingent inputs and vary in effect from situation to situation. I will call this the “changing eternity” approach. It starts from intuitions (always quite vague) about what’s “always” right and then seeks to “make them work” in real world situations. To do that, it always has recourse to something like jurisprudential reasoning. 2. Other philosophers, persuaded that eternal verities are beyond human ken, conceive norms strictly as social conventions, but, to explain the obligatoriness characteristic of norms, they typically shift attention to the fact that conventions have a purpose. Although a hammer is an artificial human construct, there’s nothing relative or subjective about the way a hammer, rather than a screwdriver, is suited to driving in nails. If conventions are instruments serving a known purpose, there will be similarly objective reasons why some are more effective than others. Such rationally-explicable facts of the matter are things experienced and thoughtful people can reasonably hope to reach agreement on. The recovery of absolute validity is accomplished–albeit in a scenario that has never been anything more than imagined—if the conventions in question are ones that aim at something everyone undeniably wants society to deliver. I call this the “utility” approach because it first asks about “what works” and only then asks about rightness, which it broaches in such terms as: how well it works, whether it works for everybody, if anything else could work better, and so on. Because it already starts from intuitions about what works in the real world, it has no need for jurisprudential reasoning. The accommodation with contingency has already taken place. It should be evident that we can put Plato, Aristotle, Augustine, Aquinas, Ockham, and Suarez in the first category and Protagoras, Epicurus, Hobbes, and Burke in the second one. The Natural Law Theory of Xunzi Without a doubt there are tensions in Xunzi between realist–absolutist and constructivist–relativist treatments of norms. In the “Discourse on Heaven,” Xunzi sounds like he is in agreement with Mengzi. Human beings become bad because they corrupt their original nature: they becloud their Heavenly lord, disorder their Heavenly faculties, abandon their Heavenly nourishment, and turn their back on their Heavenly dispositions (tr. Hutton; cf. H 176; K 3.16). This strongly suggests that the natural development of human faculties would be healthy and good. Norms are apparently an objective feature of human nature: they are operational constraints entailed by its design. But in “Human Nature Is Bad,” Xunzi pillories Mengzi and insists that human nature, left to its own natural development, produces nothing but savagery (H 248–254; K 3.150–8). Norms create social order and spare us man-on-man violence by doing violence to human nature first. But even in this pessimistic context, Xunzi returns to the theme of his “Discourse on Heaven.” Anyone can become good simply by conforming to the objectively available standards of goodness: “So, then, rén, yì, fă, and zhèng have an orderly system of principles (lĭ) that can be known and applied.”34 How does that square with standards not being naturally at our disposal, with their being artifacts manufactured to destroy or suppress what occurs sua sponte? 然則仁義法正有可知可能之理 (CTP, Xunzi 23.18), translation mine; cf. H 254, K 3.158. All Chinese passages are from the Chinese Text Project (https://ctext.org/xunzi). As a very tentative glossary, I would propose: rén – caring behavior towards others resulting from properly cultivated affect; yì – fitting behavior towards others that is differentiated by a correct appreciation of the significance you, in whatever role you play in their lives, have for someone in the kind of role they play in your life; fă – rules of acceptable conduct that apply to people irrespective of their specific roles; zhèng – correct application of fă to individual circumstance, or a correction of (including punishment for) instances of deviation from fă. 34 34 In his pronounced ambivalence, Xunzi fits so many patterns I have found looking at western natural law theory that it would be foolish not to see what we can learn from the similarities. I group them according to four ways I have shown natural law theory to be paradoxical: Changing Eternity, Common Self-Interest, Constructive Realism, and Absolute Convention. We will see that Xunzi not only flirts with all the same paradoxes as his western counterparts, but does so for the same reasons. CHANGING ETERNITY: PLATO, ARISTOTLE, AQUINAS, OCKHAM, AND SUAREZ I have documented no less than five examples of figures in the mainstream tradition of pre-modern natural law theory giving “all of the above” answers to our meta-ethical questions. In each case, their ambivalent answers reflected conflicting demands. The universal validity of norms (their necessity) requires them to bottom out in eternal, unchanging principles. But the truth of norms (what behavior they actually entail) requires them to be interpreted differently in light of each and every circumstance. Xunzi’s ambivalence seems to reflect the same conflicting demands:  Xunzi explicitly seeks non-contingent justifications for contingent norms, and non-contingent for him most certainly means “eternal.” Two of his favorite words are necessary (必) and always (常). In “On Honor and Disgrace,” for example, he sets out the system of eternal (常) relationships among honor, disgrace, benefit, yì, safety, and danger: preferring benefit to yì always results in disgrace and danger, etc. (H 25; K 1.189). That there is one truth for all, and one truth for both past and present, is a refrain (e.g., H 36, 224–7; K 1.207, 3.100–2). He’s never satisfied until he has grounded his argument on a principle that has no exceptions: all desires aim at satisfaction without limit (H 243; K 3.135); although human nature can be controlled, it never changes (H 27; K 1.191); appearances to the contrary, life is always desired (H 244; K 3.135), etc.  On the other hand, Xunzi insists that norms must be differently interpreted depending on the boundary conditions of their different applications: “In the exercise of regulation (zhì), one who is a (true) king first takes cognizance of the conditions and circumstances and then establishes (zhì) the implements and practices…. If the system (zhì) of the soil, region, and topography is not the same, the preparation and finishing of the implements and practices must be different.”35 In the opening pages of “Nothing Improper,” Xunzi extolls the responsiveness of the gentleman: “He forms a triad with Heaven and Earth…. Using yì to change his responses, he knows how to counter what’s deviant and straighten out the underlying state of affairs. The Odes say: ‘[It] goes left, [he] goes left, the gentleman makes it right (yí); [it] goes right, [he] goes right, the gentleman’s got it.’ This says how the gentleman is qualified to use yì to bend things straight, changing in response to the situation.”36 Xunzi evidently has as complex an understanding of the relation between eternal norms and the real situations they apply to as his western counterparts. The idea I documented in Scholastic natural law theory that eternal, unchanging standards entail different applications in different circumstances, and changing applications in changing circumstances, seems to be what Xunzi is getting at in his explanation 彼王者之制也，視形埶而制械用…. 土地刑制不同者，械用、備飾不可不異也 (CTP, Xunzi 18.14), translation mine; cf. H 189–190, K 3.38–9. 35 參於天地 … 以義變應，知當曲直故也。《詩》曰：「左之左之，君子宜之；右之右之，君子有之。」此言君子能以 義屈信變應故也 (CTP, Xunzi 3.5), translation mine. Classical Chinese does not require verbs to have subjects. The quotation from the Odes lacks the pronouns I have supplied. Given the context, I believe the sense of the passage is clear; cf. H 17–8, K 1.175–6. 36 35 of the dào: “This Way (sc. the true one) embodies constancy but entails (jìn: ‘reaches its extremity in’) change” (夫道者體常而盡變; CTP, Xunzi 21.5; translation mine; cf. H 227; K 3.103). To see what I mean, I think the best analogy—because Xunzi himself alludes to it—would be the basic law of optics. Implicitly everyone knows that the apparent size of objects varies inversely with our distance from them. That relation of covariation does not itself change, but it dictates how the appearances of things do change. The dào of vision is a uniform function that correlates real changes (differences in your distance) with apparent changes (differences in size). Xunzi proves fully conversant with this idea when he talks about optics and unclouded common sense. People understand that large objects will appear small from a great distance, that moving away always entails apparent change, and that proximity is the limit at which things reveal their true dimensions (H 232; K 3.108–9). A perfect illustration of Xunzi’s dào as a law of covariation is his teaching about the power of the sage. The sage does not change his or other people’s underlying nature; he transforms how it manifests by manipulating the conditions upon which its appearance depends. Human nature is eternally what it is. Desires are always without limit. But acting on desires is a matter of the heart’s approval. What economists call the “effective demand for goods” changes not just in response to changes in supply, but also in response to changes in the heart. That, too, is a uniform function. Changing what the heart approves (e.g., by setting a particularly striking example for those under you) changes what desires people consider actionable. Therefore, without changing the underlying reality of human desire, it changes what publicly manifests as desire, and it does so in a perfectly law-like way (e.g., the further from “acceptable” some behavior is, the less frequent it will be). Just so, distance changes the apparent size of objects without changing their real dimensions. Another illustration is how Xunzi thinks the Sage Kings regulated not only the effective demand for goods, but also the supply of goods. Here, instead of changing what is effectively desired (by changing the criteria according to which people filter their desires), the Sage changes what is effectively desirable (by changing what things already fitting the criteria become available). Here’s what I mean. The Sage already knows what things are naturally objects of desire, even if they are—for whatever contingent reasons—not actually (“effectively”) objects of desire. If the Sage can change social arrangements in such a way as to make what’s naturally desirable more obtainable, that marginal difference transforms what things are effectively objects of desire. But note what makes that difference important. It won’t matter how slight or fragile or contrived the initial change in the state of affairs is. Since it results in a more desirable outcome, whatever change the Sage brings about will in turn become an object of desire. That means it will tend to preserve and augment itself without further intervention by the Sage. The example to consider is the obvious one: chaos makes life less effectively desirable. At the extreme, people will decide the life their circumstances make possible is not a life they want (or conversely: the life they naturally want is unobtainable). But just as proximity allows visible objects to appear in their true dimensions, social order allows the natural desirability of human life to manifest again. That motivates people to preserve and augment the social order that makes their lives appear to be worth living. Even small improvements will suggest a gradient—a path along which life seems more worth living. In good Confucian fashion, intrinsic motivation does the heavy lifting, and, as the functioning of society gradually improves, prosperity (“increased production”) is the result. The key thing is that this result—in the absence of any coercion—is a necessary outcome. 36 The examples just considered are familiar ones—they are core Confucian doctrine. The second example is more complex than the first because we’ve added a feedback loop. But we are still operating with a law of covariation. In response to changed circumstances, human behavior changes accordingly. The functional relation itself remains constant. Behavior changes the way it does precisely because human nature doesn’t change at all. Of course, the Sage must be as cognizant of the boundary conditions as he is of the law of human nature. What it will take in any particular case to begin the “restoration of proximity” (the desirability of living) depends on how “far away” (how distressed by disorder) people were to begin with. So the Sage’s norms, although absolute, will always be applied in ways that are contingent on particular circumstances. Let’s gather what we have learned so far. We began this section by comparing two sets of passages.37 Based on these passages, Xunzi’s approach seems to have the same components as that of his Scholastic counterparts. The norm is absolute, but its relevance is relative. The norm is real, but its reality is inseparable from the specificity of constructing its application in contingent circumstances. It sounds paradoxical, but we’ve already found a convenient way to model what Xunzi has in mind. We only need to think of the eternal norm as a uniform function that takes contingent circumstances as arguments and outputs different values depending on the inputs. In the western tradition, Aquinas seems especially close to such a view. On this interpretation, it’s straightforward what Xunzi means when he says, “The Way embodies constancy but entails change.” The interpretation is based on examples that enjoy canonical status in the Xunzi and in the Confucianism generally. They leave little doubt that Xunzi sometimes has exactly this model in mind: his dào is a law of covariation. But is this the whole story? If there’s a parallel between Xunzi and the western tradition, the pressing question is how far it goes. Is it limited to the idea we can extract from Aquinas of eternal law as a stable function operating on a contingent domain? We could debate whether this is the whole story even for Aquinas. But in the natural law theories of Ockham and Suarez there’s no question. They are explicit that a simple “covariation” interpretation is inadequate.38 The “eternal function” itself is not quite uniform. It can deform depending on the novelty of the inputs. For Ockham and Suarez, things are just as Aristotle proclaimed. In matters of good and bad, right and wrong, there is nothing that is absolutely uniform—nothing that can be counted on a priori to have the same normative validity, always and everywhere. It doesn’t matter how general or abstract the formulation is.39 So the question is where Xunzi stands on this. Is he satisfied with the more abstract model of a rigid function, or does he see the need for the historicized and particularized model that Ockham and Suarez insist on? As we might expect by now, Xunzi is not consistent. He seems to lean both ways, without resolving or even acknowledging the conflict. We’ve already seen how well some of Xunzi’s most characteristically Confucian ideas fit the uniform function model. Let’s now consider passages where Xunzi seems to lean hard toward the position of Aristotle, Ockham, and Suarez. Most dramatic is Xunzi’s occasional willingness to relativize norms not just in their application, but also in their derivation. More often than not, Xunzi’s goal is to persuade us what behavior is going to be 37 See the paired bullets on p. 34 above. 38 See the exhibits on Ockham and Suarez, pp. 17–22 above. According to Aristotle (see Eth. Nic., Bks. 1–3), the only universal truth about good and bad is the vacuous one that people desire what, under the circumstances, appears good to them (1114a31–32: πάντες ἐφίενται τοῦ φαινομένου ἀγαθοῦ). What is good differs relative to circumstances (1104a), and what appears good differs relative to the kind of person you are (1114b). See my exhibit on Aristotle, pp. 16–17 above. 39 37 correct in various situations on the grounds that it is entailed by a general norm (or by the “proper application” of one). Although that remains the motivating context for the following passage, Xunzi minces no words in stating that the norm itself is first derived from the same situations to which we later apply it: Thus, to be good at discussing antiquity, one must have a measure from the present; to be good at discussing Heaven, one must enlist what is human. If one values a discourse, it’s always because it makes distinctions and classifications that match up with experience. 40 This suggest that the relativity of norms is not just the simple kind of uniform dependence of an output on a contingent input that I described above. The mapping function itself is affected if its very discovery depends on the changing value of “the present.” That Xunzi thinks the same norm he’s sure is eternal is also constructed by the sage—relative to himself, no less—could not be more evident than in the following unexpected declaration: By what means is the Sage incapable of being deceived? I say: the sage is the one who uses himself as the measure. Thus, he uses human to measure human [or: one man to measure the others], he uses [his] feelings to measure [their] feelings, he uses [his] grouping to measure [their] grouping, he uses explanations to measure outcomes, he uses the Way to determine the extremes (jìn: “the limits”). Old and new are the same. The types of things stay consistent, so even what’s ancient has the same system of principles (lĭ); therefore even if his locality is among the wicked and the deviant, the Sage will not be taken in.41 This passage is astonishing twice over. First, we could not ask for a stronger declaration of relativism. It seems it is not external standards that define the Sage, but the Sage who defines the standards. Second, we can’t but marvel how this declaration turns around on itself and becomes a justification for universal objective norms. So how does Xunzi’s dào “embody constancy but entail change”? As I showed above, it’s clear that Xunzi very often thinks of his dào as a uniform function with variable inputs. The many obvious analogies between his discussions of optics and politics leave no doubt about it. However, my purpose in quoting the last two passages has been to demonstrate that this simple interpretative solution will not always work. In passages like these, it appears that Xunzi thinks his absolutism is somehow consistent with— perhaps even grounded on—a relativity that, like the famous turtles, goes “all the way down.” Sagehood may be defined by conformity to standards supposedly set by Heaven and antiquity, but the actual content of those standards is evidently decided by a human reality that is here and now. If I have been successful in documenting an extreme tension in Xunzi between absolute and relative, real and constructed—a tension that even the “uniform function/contingent inputs” model will not quite resolve—then the question now is whether Xunzi is simply contradicting himself, impossibly trying to have it both ways, or is engaged in something more profound. For the Scholastic philosophers we considered, particularism about the world of action entailed a need for jurisprudential reasoning that effectively renegotiated the import of principles every time they were applied. Absolute and objective turned out to mean “constructed in just the right way relative to this 故善言古者，必有節於今；善言天者，必有徵於人。凡論者貴其有辨合，有符驗 (CTP, Xunzi 23.13), translation mine; cf. H 252, K 3.156. 40 聖人何以不可欺？曰：聖人者，以己度者也。故以人度人，以情度情，以類度類，以說度功，以道觀盡，古今一也。 類不悖，雖久同理，故鄉乎邪曲而不迷 (CTP, Xunzi 5.7), translation mine; cf. H 36, K 1.207. 41 38 one particular situation.” That’s surely a case of “all of the above” that goes “all the way down.” Whether it’s a coherent objective—or predicated on a latent fallacy—is harder to say. I continue to bracket that question for now. Coherent or not, there is evidence of just this kind of “jurisprudential particularism” in Xunzi’s insistence that the rules cannot be properly applied without the right person to do it: Principles alone determine nothing; categories do not apply themselves. The person who carries them (principles and categories) through survives, the one who falls short perishes. Principles are the basis of orderly administration, but the gentleman is the source of the principles. So if there is a gentleman, then even if the principles are sparse, he suffices for them to be universally effective. Without the gentleman, even if principles are implemented, there will be failures in the orderly process of execution, an inability to respond to changes in the situation, which is enough to result in chaos.42 If the proper application of norms always requires an in situ judiciousness that cannot itself be expressed as a general rule, then it goes without saying the Sage must—at least in part—use himself as a measure. Admittedly, whether it’s Aristotle or Xunzi speaking, the claim that the final standard for right and wrong is rationally inexplicable beyond the attribution that it is “what the wise man decides” leaves us wondering what criteria we use to figure out who the wise man is—especially when there are rival claimants to that position. But now we are in danger of losing sight of the solid point both of them make. We need to proceed cautiously. For the sake of argument, let’s say we can determine in advance who the wise man is. The point of jurisprudential particularism is that we will still not be able to determine in advance what the wise man will say about any particular situation. That doesn’t mean he knows something ineffable or mysterious. He will be able to explain himself to us after the fact (presumably that’s how we know he’s wise), but even he can’t know what to say about a situation before he’s actually in it. It’s the advance understanding that’s impossible. No rational ars combinatoria could ever be sure of capturing (as a “type”) the specificity of a concrete situation before it actually happens. However puzzling this situational epistemology may be theoretically, what needs to be stressed is the fact that no one finds it puzzling in practice. We all acknowledge having to rely on lawyers, judges, and juries to test and ascertain the relevance of positive law to particular cases. In other words, we all see law as presumptive until its applicability is worked out in light of the details that come to our attention only when we confront real situations. The necessary determinations can only arise case by case, as actual cases present themselves. Yes, the law itself must be something (have some unity of import) over and above its applications. But—in a workaday way we are all perfectly comfortable with—we determine what that import is only in light of its applications. According to the Scholastic philosophers we considered, the same thing is true for natural law. That means natural law acquires its objective and absolute value only when constructed in relation to actually-occurring contingent circumstances. Ockham and Suarez say as much; Aristotle and Aquinas imply it. If the concept of jurisprudence entailed by this practice isn’t self-contradictory, then Xunzi’s parallel version of “changing eternity” isn’t either: antiquity cannot operate as the measure for present 42 …法不能獨立，類不能自行；得其人則存，失其人則亡。法者、治之端也；君子者、法之原也。故有君子，則法雖 省，足以遍矣；無君子，則法雖具，失先後之施，不能應事之變，足以亂矣 (CTP, Xunzi 12.1), translation mine; cf. H 117, K 2.176. 39 times before it’s interpreted by and relative to present times; Heaven cannot operate as a measure for humanity before it’s interpreted by and relative to humanity. It’s crucial to notice that Ockham and Suarez did not adopt their position as an alternative to the uniform function model, but as a consequence of it—as a clarification of what “applying eternal law to contingent circumstances” could possibly mean. That’s why I say Ockham and Suarez were only making explicit what’s already implied in Aquinas, and why I leave open the possibility that the whole approach is predicated on a fallacy. Either a rule does or doesn’t change its meaning over time—it can’t do both. As theorists, we’d have to reject this as a bald contradiction if there weren’t such striking proof that—at least in practice—none of us sees it as a contradiction. My analysis of Xunzi suggests his path was exactly parallel to that of the Scholastic natural law theorists. In the passages considered so far, we see him making the same assumptions (unchanging law must be applied to changing circumstances), confronting the same tensions (application changes the meaning of the law), and having recourse to the same solution (case-by-case reinterpretation of what the law “eternally” means). I admit there is something “fuzzy” about this reconciliation of Xunzi’s absolutist and relativist tendencies. But that’s because it’s the nature of law in a changing world to be fuzzy until jurisprudence works its situational magic. In any case, if Xunzi can be accused of courting fallacies, then so can Aristotle, Aquinas, Ockham, and Suarez. The ambivalence Wong documents is not a “Xunzi thing” or “Chinese thing” or an “ancient thing.” It’s universal. It continues today in the ordinary practice of jurisprudence. The remaining three comparisons suggest less enigmatic ways Xunzi may be embracing “all of the above” without simply contradicting himself. COMMON SELF-INTEREST: PROTAGORAS AND EPICURUS As Protagoras and Epicurus saw it, norms are literally things human beings make up in order to try them out. So of course they are constructed. Human beings do this because they want something, so of course the norms are relative. But if things work out as planned and because of the plan, it means the construction has utility that is objectively validated—it’s an instrument that works. From there, one can even recover a claim to absolute validity. If the goal being pursued is something everyone wants, then success effectively means the norm employed to achieve it is valid for everyone. That’s definitely an “all of the above” answer. We can find strong echoes of this approach in Xunzi. He is often focused on what kind of practical policy will work in achieving social harmony and prosperity (see H 23–9, 195, 246–7; K 1.186–194, 3.43–4, 3.137–8). He typically insists on the objectivity of his policy (the other proposals are wrong because they result from some kind of bias or distortion). While he delights in associating this objectivity with the derivability of the policy from a higher norm, he not infrequently associates it with nothing more than a track-record of practical success. Is our absolutist selling himself short? I doubt it. There’s no reason Xunzi need be less than content with such a utilitarian test of objectivity. Just like Protagoras, he seems to be able to take a utilitarian stance and still recover an absolute position by making what everyone wants his starting point. This, I take it, is the significance of his insistence that, despite occurrences (such as suicides) that would seem to suggest otherwise, everyone in fact values life (H 244; K 3.135). It’s no doubt important that “everyone” doesn’t just mean everyone at birth (who are all bad), but everyone period (the good and bad alike). If that principle stands, then any social arrangements that result in making life appear undesirable must be objectively wrong, and the social arrangements that result in life 40 appearing with its true value must be absolutely right. So, right here we have at least one way Xunzi can coherently reconcile “all of the above” without any hand-waving appeal to the case-by-case practice of jurisprudence (i.e., to the “gentleman” who “carries the application of principles and categories through” in each contingently changing situation43). CONSTRUCTIVE REALISM: HOBBES For Hobbes, it’s straightforward how norms can, without contradiction, be both constructed–relative and objective–absolute: there is no justice if you don’t create it, but there is only one way to create it. If Xunzi is operating with a similar understanding, reconciling his emphatic commitment to constructivism (H 248–254; K 3.150–8) with his emphatic commitment to realism and absolutism (H 176–7, 233; K 3.14–6, 3.110) may turn out to be easier than expected. What Xunzi says about the gentleman forming a Triad with Heaven and Earth (H 176; K 3.16) makes this look promising. Let’s think first about what Xunzi does not say. He does not say Humanity is a Monad. If Xunzi were a relativist and constructivist, pure and simple, he would say that Humanity is a Monad, not one element of three that must align itself with the other two. On the other hand, Xunzi does not talk of Heaven and Earth as a Dyad to which human beings simply have to conform. If Xunzi were a realist and absolutist, pure and simple, that’s what he would say. Being the third element of a Triad means both adding something to the Dyad and deriving something from it. The gentleman must make himself a measure, as Xunzi says in “Against Physiognomy” (quoted above; see footnote 41), but not without “nourishing himself on a resource that is not of his own kind,” as he says in the “Discourse on Heaven” (tr. Hutton; cf. H 176; K 3.16). I note that the very idea of the Triad seems designed to break the dichotomy presupposed by the question we’ve been asking: “Are social norms fabricated de novo by human beings, or do they germinally pre-exist in a humanity-independent reality (in ‘Heaven and Earth’)?” Xunzi’s answer to that either–or question is simply: yes. Chris Fraser has noted that the word for Triad (參, cān) can refer to a practice for determining compass points with three stakes: the first stake would represent the Earth, the second, placed anywhere on the line between the first stake and the position of the rising sun on the horizon, would represent Heaven; the third, placed on the same line as the first two and at the same distance as the second, but in the opposite direction, would represent Humanity.44 If the resulting line is made the diagonal of a square, its four corners mark the four compass directions.45 Compass directions are norms that guide the behavior of travelers. They make a good metaphor for social norms. Notice that these orientation guides (compass directions) cannot be “found” simply by looking for them. They would not exist if we did not go through the process I just described of constructing them. Nevertheless, the outcome of the construction is in no way arbitrary. Notice, also, that the nonarbitrariness of the outcome depends on steps that are themselves purely conventional. It doesn’t matter where we plant the first stake, or how far away we plant the second one, but once we plant the second one we must adhere to that precedent, planting the third at the same distance in the opposite 43 See footnote 42 above. Chris Fraser, “Language and Logic in the Xunzi” (2016); fn. 15. In: Hutton (ed.), The Dao Companion to the Philosophy of Xunzi (Dordrecht: Springer Science+Business Media B.V.), 2016; see p. 299. 44 45 If I understand the procedure correctly, it must be carried out on the day of the equinox. 41 direction. Notice, finally, that the line on which the “Humanity” stake must be planted is already wholly determined by the first stake and the position of the rising sun. If that’s all we needed to proceed, the norm for Humanity would just be a matter of conforming to the Dyad of Heaven and Earth. But the third stake can’t be planted at all unless there is an already existing convention stipulating what the proper distance is for planting the third stake. The distance of the second stake must be arbitrary for the distance of the third to be mandatory. This model makes luminously clear how norms can be “all of the above”—once again without any appeal to a mysterious power of in situ judiciousness. It also shows how apt the comparison with Hobbes is: a) there is no compass (including a moral compass) if we don’t create it (for Hobbes: no social norms if we don’t create them), b) we can’t create a compass at all without introducing purely conventional elements (for Hobbes: no norms without covenants), c) those conventional elements are useful only if they are binding (for Hobbes: no valid covenants without an all-powerful sovereign to enforce them), and d) there is only one way to establish orientation because, however different techniques may initially seem to be (due to their conventional content), they must, in the end, align with one and the same set of humanity-independent features of Heaven and Earth (for Hobbes: whether there is war or peace is an objective fact, and there is only one kind of arrangement that successfully orients behavior toward peace). There is plenty of good sense to be found in this workaday model of ethical norms. We know that “orientation” is something more than the bare reality of reference points because we can’t fix such points and orient ourselves to them without first inventing certain conventions that make the heuristics possible. The conventions are groundless and justified only in retrospect by their utility in enabling us to fix points of reference. On the other hand, we know that “orientation” is not just a convention—that the points it identifies do exist independently of our methods for finding them, even if “orientation” does not. ABSOLUTE CONVENTION: BURKE Edmund Burke insists that norms are constructs contingent on human history, but he also insists that tradition has produced the one uniquely possible correct construction. Xunzi is notorious for maintaining the same odd convergence of beliefs as Edmund Burke: contingent tradition created our norms (H 201, 248–251; K 3.55, 3.150–4), but those norms are uniquely and irrefragably right (H 29–31, 181–2; K 1.194–5, 3.21–2). We know how Burke reconciles these views. Nothing tops the actual historical genesis of social practices for delivering the best result possible under the circumstances. Social norms are like a leather shoe: if you wear it long enough, it will fit your foot perfectly. The way Xunzi understands the achievements of the Sage Kings suggests he might have something similar in mind. 42 Michael Puett has pointed out that Xunzi’s model for the Sage Kings is not the cultural heroes who invented new tools, but the ones who perfected skilled practices, such as archery.46 The following passage leaves little room for doubt: Chui invented the bow; Fu You invented the arrow. But Yi comprehended (jīng: “achieved comprehensive finesse at”) shooting…. From antiquity to now, it’s never happened that there was duality and yet an ability to comprehend (jīng) it.47 It’s implied that Chui and Fu You were unable to equal Yi because they were divided. Why? If you’re making bows and arrows, you can’t possibly make either without thinking about both: “bow” and “arrow” are strict correlatives. So there’s no way to fabricate something like a bow or an arrow without being “divided.” But when it comes to the relationship between these two fabrications—for Xunzi, that’s another thing altogether. The relation between them is a unity. Given the way Chui and Fu You devised the bow and arrow, Xunzi evidently believes there is only one optimal “fit”—only one way to relate them in the undivided practice of hitting a target. That’s what he thinks Yi discovered and bequeathed to Chinese society. Both the unique validity of traditional norms that Xunzi insists on and the contingency that he concedes follow from this understanding. If, for the kind of bows and arrows in question, Yi’s technique achieved maximal success, how could it be improved on? It would be perverse to question its validity. But it goes without saying that anyone using a different kind of bow and arrow (ones made of different materials, for example) will need a different technique. Xunzi gives the same analysis of social life. He knows that role concepts are all correlatives: you can think of “man” without necessarily thinking of “boy,” but you can’t think of “father” without thinking of “child,” and vice versa. The concepts are related internally. To call someone “father” is to recognize not one, but two social roles. According to Xunzi, human beings differ from lower animals precisely because they grasp such relationships: “The birds and beasts have fathers and sons, but they lack the intimate relationship of father to son. They have male and female sex, but lack the man-woman distinction.”48 Xunzi is saying that human relationships are both socially constructed and objectively real, and he is making a valid point. The unity of a relationship (e.g., “brotherhood”) is nothing tangible. It exists only because human beings construct it in thought. But it is not an arbitrary fiction: first, because things really are related in one way or another (brothers are tangible things, even if brotherhood is not) and, second, because whatever the specificity of the relationship-conception adds that wasn’t “real” to begin with does have real consequences (more or less desirable ones) when people act in conformity to it. Purely conventional relationships creating such roles as “supervisor” or “guarantor” are perfect examples—the consequences of having a “guarantor” can be quite tangible. Loading natural relations with conventional content (as we clearly do, for example, with “brotherhood”) has the same effect. A society in which something intangible like “brotherhood” is treated as a thing (a thing to be cherished or cultivated, for example) is a society in which constructed norms have a basis in reality (a biological basis). If, moreover, the construction has the happy effect of promoting human well-being in tangible 46 Michael Puett, op. cit., p.70. 倕作弓，浮游作矢，而羿精於射；… 自古及今，未嘗有兩而能精者也 (CTP, Xunzi 21.10), translation mine; cf. H 231, K 3.107. 47 48 夫禽獸有父子，而無父子之親，有牝牡而無男女之別 (CTP, Xunzi 5.5), translation mine; cf. H 35, K 1.206. 43 ways, the construction can—for the kind of reasons Burke was good at making clear—lay some claim to objective reality. But what about absoluteness? Surely such constructed norms are relative, even if their effectiveness is “cum fundamento in re” (based in reality). Wouldn’t other constructions be possible that were just as good, rendering any preference for one in particular purely relative? Xunzi’s insistence on the unique rightness of his tradition tells us most of what we want to know. Xunzi evidently thinks relating social correlatives in an “undivided practice” is exactly analogous to archery, as he understands it: there is a target to be hit, and, given the instruments provided, only one optimal technique for hitting it. If the technique passed down to us from the Sage Kings succeeds in “hitting the target,” it would be perverse to think we could get equally good results with a different technique. Either the shoe fits or it doesn’t. If the evolved practice “fits,” what more is there to consider? It’s obvious that Xunzi’s uniqueness claims are wrong—both for archery and society. Perhaps he couldn’t imagine how different techniques using the same instruments could get equally good results. But it’s also possible he realized you can’t get anywhere unless you put the first stake down somewhere and the second stake at an arbitrary distance you accept as binding. Either way, it seems Xunzi can embrace “all of the above” without simply contradicting himself. And once again—in case one finds it problematic— no appeal to anything like jurisprudence is required. RÉSUMÉ OF WESTERN NATURAL LAW PATTERNS IN XUNZI’S THINKING Xunzi’s dào looks a lot like western natural law. Just like natural law in western theory, Xunzi’s norms are both absolute and relative, both objectively real and constructed. And just as with the western theories, in Xunzi we find two distinct strategies for reconciling these opposites. One (the changing eternity approach) stipulates eternity and universality for principles, but then finds a way to accommodate them to change and contingency. The other (the remaining three, which are variants of a “utility” approach) stipulates the utility of some contingent practice, but then finds a way to deem it objective and necessary. Xunzi’s “changing eternity” approach is no more incoherent than the familiar western concept of jurisprudence. I will grant: how jurisprudence actually operates is somewhat mysterious and may be impossible to state theoretically. After all, if one could state it theoretically, the practice itself would be unnecessary.49 Does that mean the practice is logically suspect? I don’t claim to know. But Xunzi is no worse off than his western counterparts. The other three approaches have the virtue that they can be spelled out theoretically, and I believe I have shown that each of them is logically coherent: 1) Common Self-Interest: Practical success in achieving goals everyone shares establishes the means as an objective norm that is, for all intents and purposes, universal. 2) Constructive Realism: Norms have to be constructed because there is no way to successfully align behavior with real reference points in the world without introducing pure conventions. 49 Just as no axiomatized system will ever be able to prove all truths in its own domain of discourse, no theory of jurisprudence will ever be able to generate or compass all the principles that jurisprudence will employ to reach valid decisions. 44 3) Absolute Conventions: Practices that invest real relations with conventional content have real effects, and some of these networks of conventional relations (“norms”) are successful (some may even be uniquely successful) in promoting the well-being of the network’s members. It is beyond my scope to work through the details here, but all of these approaches suggest ways the overt constructivism and relativism of “Human Nature Is Bad” could be logically consistent with the overt realism and absolutism found in the “Discourse on Heaven.” It may even be the case that Xunzi is combining all three of the “utility” approaches. In the third utility approach, for example, it may be that the specificity of the goals embedded in the way the Sage Kings defined social relationships (i.e., what counts as “hitting the target” in the “undivided practice” of, say, being “father and son”) is, in the final analysis, nothing but an arbitrary convention. But if so, it may be arbitrary in the way the second utility approach requires: as a necessary precondition of there being something mandatory. Far from eliminating objectivity, it’s what would makes the rational calibration of benefits possible. Now the first utility approach would fit in, too. Once goals are specified, how well the stipulated goals are being met by the means prescribed can be measured—it’s an objective question of success and failure. If observing the required conventions brings the desired tangible results, they are objectively validated. How this integration of the three utility approaches might fit with the “jurisprudential” approach is a matter for further research. Conclusion Is natural law theory logically incoherent? I have vacillated, speaking now of “contradiction” and “having it both ways,” now of “tension,” “ambivalence,” or “paradox.” If I had to choose, I am more open to affirming natural law is self-contradictory than to letting go the idea that its life and soul is a seemingly impossible act on a tightrope. But I don’t think what it tries to do on the tightrope is impossible. The Xunzi is a perfect object lesson. There’s good reason why exacting, scholarly interpretations of Xunzi’s meta-ethics run the gamut—as David Wong has documented—from relativist and constructivist to absolutist and realist: each of these readings finds its share of undeniable textual support. But that very fact proves none of them is right. If Xunzi is not simply contradicting himself (in which case he has no position), he must be occupying a middle position that frustrates our dichotomies. If he’s an absolutist or a realist, his position is nevertheless softened by a recognition that the same standard needs to be applied differently in different circumstances. If he’s a relativist or constructivist, his position is nevertheless qualified by recognition of objectively real limits on what will work. There’s a reason I use the phrase “all of the above” and not “each of the above.” If Xunzi doesn’t hold any of the above positions, then he certainly doesn’t hold each of them—that would be a contradiction. But there’s a way he can hold all of them by qualifying each. What I have just said of Xunzi applies, as far as I can tell, to every philosopher in the western natural law tradition. Those, such as Protagoras, Epicurus, or Hobbes, who declare themselves relativists or constructivists, are not subjectivists or skeptics. Just like the official utilitarians, who claim to revile natural law theory, these unofficial utilitarians focus single-mindedly on the objective conditions of successful action. The conditions of success are of course different in different situations, but they are the same for everyone in the same situation—and the same for everybody, period, when everybody is in the same situation and wants the same thing. So, without reneging on their declared starting point, the avowed relativists and constructivists end up adopting positions thought to be characteristic of their 45 adversaries. That manoeuver requires a distinctive way of getting there, but it’s unquestionably realism and absolutism they sidle up to. Their adversaries have a complementary manoeuver, approaching the same omnibus position from the opposite direction. Contrary to how they are usually represented, those who declare themselves absolutists and realists (I have documented this for Plato, Aristotle, Aquinas, Ockham, and Suarez) are not authoritarians purveying a one-size fits all dogmatism of procrustean absolutes (a compact set of codified rules presumed to be valid for everyone, always and everywhere). To be sure, there were supposed to be some natural laws that were “primary,” and these primary natural laws were supposed to be “eternal” and have imperative force. But that didn’t mean we could forgo all construction relative to contingently changing circumstances and still talk about what natural law requires us to do. It certainly didn’t mean we could, in abstraction from all application, itemize “primary” natural laws once and for all, producing a self-sufficing “short list.” That’s what Grotius, Hobbes, and Pufendorf tried to do. For the unabashed realists and absolutists of the pre-modern tradition, the very universality of the primary natural laws meant they could not be anything but more-or-less vague axiological priorities until interpreted in light of real contingencies. Yes, they were rules of conduct, but only indirectly. Primarily they were controls on the process of jurisprudential reasoning. To bear directly on conduct, they had to be loaded with contingent information about ever-changing circumstances. So, without reneging on their declared starting point, avowed absolutists and realists also end up adopting positions thought to be characteristic of their adversaries. I know this last claim is the one that will meet the most resistance. I am contradicting a view that conventional wisdom feels especially confident about. Whether celebrated or reviled for it, either way the pre-modern natural law tradition is known for its unreserved commitment to realism and absolutism. What else could that commitment mean but that its natural laws were supposed to be an articulated set of discrete permissions, commands, and prohibitions that were always valid and could be inscribed once and for all—just like the Decalogue or the Twelve Tables? But it’s demonstrable the premodern tradition never thought of natural law that way. It can’t be stressed enough that Hobbes was the first person to generate an itemized list of purported natural laws (nineteen in all). His inspiration was Grotius, who, in the Preface to his The Rights of War and Peace, was the first to even toy with the possibility of some kind of “short list” enumeration. The fact is damning to the conventional view: no Roman jurisconsult, no civilian or canon lawyer in the Middle Ages, and no pre-modern philosopher, whether ancient or medieval, ever tried to catalog the natural laws. If they thought natural laws were the sort of thing that could be cataloged, the legal thinkers of the great pre-modern traditions would have seen it as their professional duty to do just that—and they would have done it with the same gusto the rationalists of the 17th and 18th centuries devoted to that project. But from the time of the jurisconsults up to the Renaissance, it was common only to illustrate what sort of thing natural laws were by giving a few vague examples. It was a pitifully small handful—always the same canonical few, such as honoring parents and promises, that originated with Cicero and the early jurisconsults. What to conclude from their perfunctory repetition? One thing is clear: if a complete inventory of explicitly defined fundamental rules was never attempted, it can only be because the possibility of such a thing was never imagined. But that means our conventional wisdom is wrong: pre-modern legal theorists did not think of natural laws as a set of commands and prohibitions that could be definitively articulated.50 50 Some readers may wonder why the Decalogue itself is not an obvious counter-example to my argument. Was it not identified by many pre-modern philosophers with natural law? This is a tricky question, but I think the answer is no. Even for pre-modern 46 For sure, from Augustine to Suarez, the “eternity” of natural law was invoked. It was invoked because jurisprudential reasoning aspired to bottoming out in evident principles of exceptionless generality. And, yes, that meant: in unchanging principles that were eternally applicable. So much, so good! The question is: did discovering such principles mean they could be codified? “Applicable” means applicable to a changing world. Was the “eternal” part of “eternally applicable” something that could be factored out and separately displayed, apart from any applications? Until the rationalist idea of a priori construction stormed early modern Europe, no one thought so. For the pre-modern tradition, what could be codified was, by definition, human law. If just, human law was an application of eternal law, and its codifiable principles were an historically and culturally appropriate expression of eternal law. But the eternal principles themselves were another matter. Eternal principles cannot be codified.51 By their very nature, they are too general, admitting countless articulations—that’s what makes them forever useful in the face of unanticipated applications. At the same time, however, the objective absoluteness of natural law was never questioned. Precisely because of that, the articulation of natural law pushed in the opposite direction—away from the presumed but elusive eternal coding and towards its concrete application. In their purity as unchanging principles, unapplied natural laws were simply not capable of any commensurate expression. They could only be hinted at with boiler-plate examples as vague as they were familiar. The articulate expression of natural law had to move in the other direction and accommodate itself to the requirements of the intended application. Ultimately, it coincided with the one eternally right interpretation of its import in each situation to which it is applied. In the long haul from Aristotle to the late Scholastics, we saw this philosophers whose accounts of natural law explicitly referenced the Decalogue, that famous table of categorical commandments did not constitute natural law as such. They may say it does, but their own practice belies their words. They treated it as an historically revealed formulation of law. No one doubted it contained high-level principles of natural law, but these had to be extracted by a process of interpretation from the positive formulations recorded in the Pentateuch. This extraction was not a matter of revelation and not an exercise in pure reason. As my exhibits make clear, it could only take place as part of the larger process of interpretation in light of concrete applications. In other words, pre-modern philosophers treated the eternally right injunctions they supposed the Decalogue to contain not as definitive rules of human conduct, but as highlevel controls that must be observed if our reasoning about norms is to count as “natural.” Crucially, the controls themselves were treated as adequately specifiable only in and through the process of trying to observe them. Ockham and Suarez say as much, Augustine and Aquinas imply it. Of course, the natural law theorists typically equivocated and called both the principles and their specific applications “natural laws.” But that’s because they understood natural law to encompass the whole process of legal reasoning, which moves back and forth between the determination of presumed controls and the determination of case-by-case outcomes. It was not because they thought the principles were “already” natural laws in and by themselves and so specifiable apart from the process of jurisprudential application. Because natural laws could not be isolated in that way, they also—to put a point on it—could not be carved in stone. For pre-modern natural law theory, the written codes of Exodus 20 and Deuteronomy 5 are historically contingent applications. Of course, they are singularly important applications, but they are still only applications—not raw statements of pure natural law principles. 51 I am, of course, talking only about normative principles of human action. No one I’m discussing doubted that the eternal principles of a theoretical science, such as geometry, could be codified. In this light it’s easy to see that the revolutionary innovation of the 17th century rationalists was a fairly simple trick. They simply rejected the traditional theory–practice distinction. If all knowledge is theoretical, it means even the norms of human action can be known and applied more geometrico. However, as noted above in my discussion of Pufendorf (pp. 8–9), there is a cost to making the principles clear and explicit enough to codify. Since they need to be applied to a particular social and political reality, the more clear and explicit they are, the more they must be tailored to fit that reality. They end up prefiguring—sometimes quite specifically—the social and political world to which they are being applied. Heinrich Rommen, citing Anselm Desing (see fn. 5 above), saw this as proof that something was wrong with modern natural law theory. However, as I have shown, limiting its application by conforming to contingencies is simply what natural law does. If principles are left vague, they seem universal because conformation appears only in the applications. As soon as principles are spelled out explicitly, the conformation begins to appear in the principles. Rommen is certainly right that Pufendorf only imagined his explicit system of natural law was universal. But Rommen is guilty of the same mistake (see fn. 53 below). 47 push toward the “absolute” particular again and again. For a traditional realist and absolutist, then, there’s simply no way natural law could be prefigured, once and for all, in any finite inventory. Its absoluteness can indeed be expressed, but only in relation to the specificity of its applications, which means: at the tail-end of indefinitely many complex and decidedly ad hoc processes of jurisprudential construction. In each case, of course, the conceit is that the same eternal principle has been preserved precisely by expressing its import differently. I must stress that pre-modern natural law theorists did not just shy from a putative complete catalog of natural laws (the sort of thing Pufendorf aspired to). They didn’t even attempt to catalog a short list of highest first principles the way Grotius did.52 That proves they had no inkling of a priori construction. They were jurisprudential fallibilists “all the way down,” understanding that even their grasp of the highest principles was presumptive: partly dependent on the very attempt to apply them and therefore always subject to ongoing interpretation and possible—indeed, inevitable—correction. Pre-modern theorists certainly thought the principles of natural law were first in the “order of nature,” but both their theory and their practice prove they held them to be last in the “order of inquiry.” Crucially, that was an ever-developing inquiry they knew could never be completed. So we arrive very clearly at the point I’ve hoped to clinch all along: even the highest “eternal verities” of the traditional realist and absolutist—not just the specific “applications” of them—are exceedingly hard to distinguish from the relativities of the historicist.53 Notice, please, that I am not simply reclaiming traditional pre-modern natural law for relativism. Rather, I am insisting on an “all of the above” approach in which absolute, real, and eternal cannot be cleanly separated from relative, constructed, and historical. The point is that even for a realist and absolutist, 52 It might be objected that some pre-modern legal theorists did articulate highest principles of natural law. But these alleged principles were nothing more than tautologies (such as “good is to be done, evil avoided” or “justice is giving each one his due”). These weren’t meaningful precepts, but place-holders for meaningful precepts whose actual expression depended on a determination of what was good or due. The closest anyone came to specifying what was eternally good or just was with the boiler-plate examples discussed in the text. 53 A modern neo-Thomist such as Heinrich Rommen would disagree completely with my assessment. For him, natural laws are not only discrete precepts on the model of the Decalogue—they are the Decalogue (op. cit., pp. 45–51, 55–61; it goes without saying that we are talking about a highly interpreted Decalogue that already meets the needs of philosophical rationalization). I will briefly address why I think Rommen has no counter-argument against the interpretation advanced in this paper. Like all neo-Thomists, Rommen believes he is faithfully following St. Thomas (whose rational interpretation of the Decalogue he simply takes for granted). But, as we have seen, the “eternal absoluteness” of natural laws—even when they are being associated with the Decalogue—is only half the story for any Scholastic thinker. Rommen, in fact, fearlessly follows St. Thomas in stressing the other half of the story—but he doesn’t seem to grasp the implications of doing so. He explicitly admits that natural laws must be interpreted by positive laws in order to be applied (pp. 58ff.). But that’s tantamount to admitting they are not really absolute—that their very abstractness prevents it. In other words, without realizing it, he concedes that natural law must always be partly “adventitious” (cf. pp. 20ff. above) and that it must be so not despite, but because of its “eternal applicability.” How what Rommen ends up endorsing differs more than superficially from the rationalist procedures of the 18 th century that he so theatrically reviles (pp. 93ff.) is a mystery to me. We have seen that, for Scholastics, natural laws are something like guidelines for jurisprudential reasoning about specific applications and that the guidelines themselves are not eternally fixed because their meaning evolves in tandem with their applications. When neo-Thomists turn these elusive principles into explicitly articulated precepts modeled on the Decalogue, the latter become extremely hard to distinguish from the itemized principles of the early modern rationalists. But little else changes. Recall the paradox of Suarez: given the circumstantial “essence” of conduct, a definitive (“one size fits all times and places”) table of “conduct rules” is inherently impossible. Consequently, even for rationalists and rationalist-leaning Thomists, practical guidelines can have only limited intelligibility apart from what arises in and through the hermeneutics of their application. Despite the magisterial display of codification, the meaning of the principles will remain crucially dependent on their applications. Codification, in fact, makes it more explicit by prefiguring the social reality the codifier needs the principles to apply to. That’s precisely what we see with Rommen—an uncompromising activist Roman Catholic. For Rommen, the assumption that primary natural laws can be explicitly codified leads exactly where it must: to the linguistic fixation of Thomas Aquinas’ rationally-interpreted Decalogue. 48 the true “eternal” import of a natural law—its presumed inherent unity of meaning—emerges only gradually. It emerges through our repeated efforts to salvage its unity in the face of the diverse and unpredicted contingencies to which, again and again, it must apply. This process is entirely rational. It’s regulated by the ideal of justification through an eternally valid norm (where the norm logically entails the permissibility of a class of actions without generating conflicts with other norms). But—as I have insisted throughout this paper—it’s also entirely ad hoc, both stimulated and limited by the vagaries of history. After all, what consumes our efforts at rational control is nothing other than what actually happens. I have documented a surprising convergence between relativists and constructivists, on the one hand, and absolutists and realists, on the other. Is it hard to explain? Effective ethical thinking seeks to make non-arbitrary judgments about contingent situations. There’s necessarily a realist component (to make it non-arbitrary), implying some truth-of-the-matter that is generalizable because absolutely valid within specifiable limits. And there is necessarily a relative component (to make it relevant to real life), implying some contingently-motivated construction. I ask: what ethical philosopher doesn’t end up providing us with both components? How many declared relativists are dyed-in-the-wool subjectivists of the “anything goes” variety? How many declared realists are rigid fundamentalists who forbid any hermeneutics as casuistry (in the negative sense of the term)? The real payload of Wong’s analysis may be his conclusion that a constrained relativism and a flexible absolutism are, for all intents and purposes, the same position, just outfitted with different ideological labels. While initially they draw different lines in the sand and name themselves accordingly, they each make complementary concessions and end up in almost the same place. So it’s not necessarily a contradiction if Xunzi is both a “relativist” and an “absolutist,” and we should not be surprised if the same thing is true of most ethical philosophers.