Jacques Maritain Center: Ethics Without God?

Hierarchy and Direction for Choice

Daniel McInerny

Draft July25, 2003

This essay is a response to an invitation. Robert George has recently requested from "neo-scholastics" "a detailed account of how choosing in accordance with a principle of hierarchy is supposed to work across a set of cases."{1} George's invitation is put forward in a defense of what he calls "the incommensurability thesis" and its role in giving direction to human choice. Put briefly, and quoting George, "[t]he incommensurability thesis states that basic values and their particular instantiations as they figure in options for choice cannot be weighed and measured in accordance with an objective standard of comparison."{2} This raises a crucial question. In the absence of an objective standard of comparison, how is non-arbitrary choice between basic goods possible? For George, it is possible because the very respect that is due to the incommensurability of the basic goods generates principles that objectively guide human action. What reasons does George give for rejecting a natural hierarchy of human goods? First of all, it should be said that George does recognize certain natural hierarchies. Instrumental goods, he acknowledges, are subordinate to intrinsically valuable goods; sensible goods are subordinate to intelligible goods; and a fuller realization of the good is to be preferred to a meager realization.{3} When George speaks of incommensurable "basic" goods, therefore, at issue is not these hierarchies but rather the set of intrinsically valuable, intelligible goods that together comprise human happiness. Similarly for the Thomist, the argument for hierarchy principally concerns relationships between intrinsically valuable human goods.

So why then does George reject a natural hierarchy of basic goods, and the making of moral judgments on the basis of such a hierarchy? The reason he gives is put in the context of the following example. Imagine a golfer faced with a decision between interrupting his best chance to break 80 and the opportunity to rescue a child drowning in the water hazard on the adjacent fairway. Granted that the only moral choice for the golfer is to drop his club and try and rescue the child, the question is how this decision is reached. For George, what is clear is that the golfer does not make his decision by acknowledging that the basic good of "life" is objectively better than the basic good of "play." Why not? The answer is twofold: because either the notion of hierarchy places in jeopardy the common sense belief "that one ordinarily has no moral duty to forego one's ordinary pursuits, including playing golf, to devote oneself to life saving or to join famine relief projects and other worthy lifesaving endeavors in far off places," or the notion of hierarchy fails to provide a principle on the basis of which to decide when choices for "life" are required and when they are not.{4} In other words, for George, either hierarchy forces us to fanatically reduce the value of lower goods in the hierarchy in favor of higher goods, and indeed the highest good, or it simply fails to do any work in telling us when choices in favor of a super-ordinate good are required. In either case, hierarchy seems hopeless in providing direction for choice, while the incommensurability thesis allows for the intrinsic choice-worthiness of basic goods even while it grounds the principles that would direct us at times to favor one basic good over another.

George is exactly right that the incommensurability thesis runs counter to what he calls the "neo-scholastic" -- I would say principally Thomistic -- understanding that the human good exists naturally as a hierarchical arrangement -- a duplex ordo as Aquinas describes it in the first lectio of the Commentary on the Ethics -- in which goods are ordered both to one another and to the absolutely ultimate end.{5} So in responding to George's invitation, what I aim to do in this essay is to provide an introduction, at least, to a Thomistic understanding of how the natural hierarchy of human goods provides direction for choice. In sum, I will be arguing that non-arbitrary choices between contending substantial, or intrinsically valuable, goods are only possible when one of the goods is seen either as a necessary or expedient means for the attainment of another and intrinsically more valuable good. For Aquinas, in any line of action what is obligatory in itself is the intrinsically valuable good, or end, and that which is for the sake of this end, the means, is obligatory on account of it, either by being necessary or expedient to the attainment of this end.{6} In determining, then, which of two or more choices is necessary or most expedient to the attainment of an intrinsically superior end, deliberation must first discern the appropriate natural hierarchy of human goods.

Several important issues hang on whether the incommensurability or hierarchy thesis wins this debate. Most obvious is the issue of how to justify non-arbitrary choices between conflicting, substantial goods. But analogous to this is the issue of how to justify non-arbitrary choices between entire life-plans. Would I have non-arbitrary reasons for becoming either a butcher, baker, or candlestick maker? This latter issue touches upon a further question. If there is indeed a natural hierarchy of human goods, does this imply that there is one and only one "right" way of living a human life? Where in the argument for hierarchy, in other words, do considerations of wish, talent and temperament come in?

The debate between incommensurability and hierarchy also raises the question of how to adjudicate between the bonum mihi and the bonum commune, the good that is a personal good for me here and now and the good that by nature is shared with those with whom I live in community. Is there a hierarchical ordering between these two kinds of good, or are they themselves incommensurable considerations of human agents?{7}

But the most important -- because the most fundamental -- issue raised by this debate concerns the unity of the human good. For Thomists this comes down to the question of the ultimate end, of happiness. Thomists understand the ultimate end to have several analogous senses, the most primary of which identify the ultimate end with God. For the Thomist, God is the principium of the hierarchy of goods, and thus God ultimately unifies the human good by providing the ultimate direction for choice.{8} Defenders of incommensurability, by contrast, understand happiness solely in an inclusivistic sense; that is, as the name we give to that miscellany of incommensurable basic goods that give us satisfaction. While the inclusivist view does well in capturing our sense of the multiplicity of the human good, it seems to do less well in showing us how the collection of incommensurables amounts to anything more than the sum of various parts. Such a multiform view of human fulfillment has serious implications for a unified conception of the human person.{9}

I want to address these issues from a Thomistic point of view in the following way. George has asked, in particular, for a detailed account of how choice in the context of hierarchy works across a set of cases. I believe it would be a mistake, however, to get down to cases right away. For the differences between the incommensurability thesis and the hierarchy thesis are differences that occur at the most basic level of understanding of the human good, and so any constructive debate between the two theses must first occur at this level. So, after first confronting a threshold challenge that any defender of hierarchy must confront, I will then develop the nature of the "for-the-sake-of" relationship that serves as the basic ligature of hierarchical ordering. This examination of the "for-the-sake-of" relationship will quickly lead us to the central question about the nature of the ultimate end, only after which we will be in a position to look at how hierarchy goes to work in some particular cases at different levels of moral decision-making.

While this response to George's invitation may not accomplish everything that George would require of the defender of hierarchy, I trust it will provide a first rough sketch of the form an adequate response to his invitation must take.

A Threshold Challenge

In his recent book, Natural Law and Practical Rationality, a natural law defense which, like George's, reposes upon the incommensurability thesis, Mark Murphy reads both George and John Finnis as issuing the following threshold challenge to any defender of hierarchy. Any defender of hierarchy must show first either that the incommensurability thesis applied to the basic goods is false, or that incommensurability is consistent with hierarchy.{10}

So to take up the first part of the disjunction: is it the case that the defender of hierarchy must reject the incommensurability thesis as false? The answer, perhaps surprisingly, is no. The incommensurability thesis is, in fact, not false if by incommensurability we mean that intrinsically valuable goods, at least, cannot be reduced to a single genus. This is one of the points pressed by Aristotle against Plato's Form of the Good in the sixth chapter of Book I of the Nicomachean Ethics. The Platonists themselves, Aristotle says, do not postulate a single form for classes of things in which prior and posterior are found, as is the case with numbers. While there are forms for individual numbers, there is no form of number itself, because the class of numbers is comprised of various natures ordered to each other and to a first. And so it is with the good. Goodness manifests itself across the categories: there are good substances, good qualities, good relations, and so on. And these various manifestations of good, as with numbers, enjoy an order of prior and posterior. That which has goodness in itself, substance, is prior to all those other goods that manifest their goodness only in relation to substance. From these observations Aristotle concludes that just as there can be no common form of number, so there can be no common form of good.{11}

Thus it is perfectly appropriate to speak of intrinsically valuable (as opposed to instrumental) human goods as incommensurable. For human goods are not commensurable in the sense that they are merely different manifestations of a single kind of good. In this respect, they have no shared mensura. Accordingly, the hierarchical understanding of good I am defending has no truck with a commensurability thesis, or with those quasi-mathematical, maximizing strategies of practical rationality that trade on such a thesis, and which natural law theorists like George are absolutely right to condemn.

So, if the incommensurability thesis is not to be rejected, then we must affirm the other side of the disjunction, namely, that incommensurability is compatible with a hierarchical understanding of the good. Both Russell Pannier and Mark Murphy have noted that this compatibility has been implicitly recognized by theorists such as George. As Murphy puts it, there is a tu quoque rebuff to George's objection to hierarchy, for George himself holds "that each person is under a practical requirement to form a life plan that includes a subjective prioritization of the basic goods in his or her life."{12} But if the basic goods are such that they are amenable to subjective prioritization, then in principle, at least, there is nothing inconsistent in thinking that the basic goods can enjoy objective prioritization. "And if George asks," Murphy writes, "what particular requirements on choice are generated by the goods' naturally forming a hierarchy, the defender of that view can respond that it is the requirements on choice that would be generated by the goods' forming a structurally identical hierarchy through the agent's commitment."{13} But what reasons are there for thinking that intrinsically valuable human goods actually do form a natural hierarchy? I would like to address that question now, as the next stage in showing how hierarchy works in providing real direction for choice.

The "For-The-Sake-Of" Relation

Earlier I referred to Aquinas's description of the human good as a duplex ordo, a twofold order of goods both to one another and to the absolutely ultimate end. As Aquinas also says in this lectio of his Commentary, the order of goods to one another is made possible by their order to the absolutely ultimate end.{14} So in setting up the scaffolding of an objective hierarchy of goods, the existence and nature of an absolutely ultimate end must be established, as well as the ligatures that bind other less-than-ultimate ends both to it and to each other. Following both Aristotle's and Aquinas's procedure, I want to consider the ligatures first, the basic means-end structures that characterize the hierarchy of our objectives.

Permit me, first, some rather rudimentary distinctions that are nonetheless absolutely necessary for the argument to follow. Say that I have volunteered to play in a charity golf tournament. The actions that I take in preparation for my play -- practicing for the tournament, driving to the site, the play itself -- are all ends that I pursue for the sake of my overall end of benefiting some needy children. The benefit of the children serves as the final end, the term, of my action, while the ends subordinate to it serve as means to this final end. We already see that "means" and "ends" are relative terms. What makes the benefit of the children the final end of this train of action is that I would desire this end even if nothing else ever followed from it: such as public recognition of my action. All the other ends, however, at least within this train of activity, would not be pursued unless they were somehow productive of the end of helping the children. I might play golf for other reasons on other days, but I would not tune up to play in a charity tournament unless I was convinced that my participation would truly help the charity. Thus of any end we can ask whether we would pursue that end even if nothing resulted from it but the attainment of the end in question. Would I floss my teeth if that activity had no effect on dental hygiene? For most of us, I hope, the answer is no, and so we identify flossing teeth as a merely instrumental activity. Would we want to play golf apart from any usefulness it may serve? Of course we would. Playing golf is intrinsically valuable, a final end.{15} This example helps clarify that certain ends can be both instrumentally and intrinsically valuable, depending on the context in which they are pursued.

The "for-the-sake-of" relationships that hold between goods exhibit what Richard Kraut has called asymmetrical causal relations.{16} In Aristotle's well-known example, bridle-making is for the sake of riding, meaning that bridle-making helps bring riding into being and thus that riding is more desirable than bridle-making -- hence the asymmetry in the relation. Moreover, because bridle-making is for the sake of riding, riding provides the standard or norm against which bridle-making is regulated. Bridle-making takes the form it does because the art of riding, the craft to which it is subordinated, takes the form it does. This is the meaning behind Aristotle's use in this context of the word architektonikê, "master-craft." The Greek term connotes a craft that is superior, of course, but superior insofar as it is an archê, a ruler, over the others. Politics, Aristotle's politikê technê, is the master-craft of master-crafts because it dictates to all other sciences and crafts both what is to be done in these crafts and how they are to be employed for the end of politics: the common good.

We are beginning to see how hierarchy provides direction for choice, insofar as the good for the sake of which another good is pursued regulates the pursuit of the subordinate good. A clarification on this point. A higher good regulates a lower good in more than one way. Commenting on the passage from Aristotle just mentioned, Aquinas claims that politics regulates practical activities both as to whether they should be pursued and how they should be pursued. But politics regulates speculative activities only as to whether they should be pursued, either at all or by a particular person. Politics does not dictate how a speculative science should be pursued -- how in geometry, for example, conclusions should be drawn from premises. For this depends on the very nature of the subject matter of geometry.{17} So in the order of practice, at least, what Aquinas has to say about the relationship between politics and other activities seems to hold generally: lower goods are for the sake of higher goods, which in turn dictate whether and how the lower goods are to be pursued.

It is a good question whether the asymmetry in the for-the-sake-of relationship must always run in the same direction. Can x be for the sake of y, which in turn is for the sake of x? The goods of family life, on any Thomistic view, are ordered for the sake of practicing philosophy. But as a working philosopher, do I not also pursue philosophy for the sake of the goods of family life? This example quickly clarifies the point that goods are not ordered for the sake of each other in the same respect, thus ensuring that there is no true circularity in for-the-sake-of relationships. For while it is true that the goods of family life are ordered to philosophical wisdom, it is not true that philosophical wisdom is ordered, as a subordinate good, to the goods of family life. Philosophy understood as employment may be so ordered, but this is just to change the respect in which we consider philosophy. Employment understood as a mere means is always subordinated to the goods of family life, which in turn are always subordinated to the philosophical pursuit of truth.

For Aquinas, all intrinsically valuable goods exist in per se, that is necessary, relationships of the prior and posterior. Prudence is a superior virtue to fortitude, according to Aquinas, principally because it perfects that power of the soul concerned with the overall good of the agent. The hierarchy among these virtues is a necessary feature of the human good, such that a conception of fortitude not put in the service of and regulated by prudence would not be the genuine article at all. In the film The Perfect Storm the members of a fishing crew lose their lives pursuing a catch straight into the teeth of the storm of the century. True, they are down-on-their-luck fishermen, desperately in need of a good catch. But clearly it was not worth risking their lives to catch even more fish than they had already caught. The daring quality they exhibit in battling the storm is in some sense impressive, the film's marketing company may describe it as "courageous," but in the absence of prudence we can't admit that it's anything other than recklessness.{18}

We may conclude from this that the very character of intrinsically valuable goods depends upon how they are subordinated to and regulated by goods superior in the hierarchy. This is why to speak of "basic goods" -- even in their instantiations -- as discrete, incommensurable items (George employs the unfortunate metaphor of 'quanta'{19}) is not to speak of real goods at all, but only of generic, ghostly entities incapable of directing choice. Once again, it is a feature of intrinsically valuable goods (at least, of all but one of them) that they are instrumental to higher goods, and that the whether and how of their pursuit are guided by these higher goods. We simply don't fully understand the good of "life," for example, until we understand that in certain circumstances it must be sacrificed, subordinated to, the familial or political common good.

Of course, George and others attempt to get around this difficulty by invoking the notion of "integral human fulfillment," an appeal to an ultimate end that brings some definition to the basic goods and provides the ultimate direction for, and justification of, human choice. Hence it is opportune at this juncture of the argument to consider the role of the ultimate end in serving as the principium of the hierarchy of goods.

Before that, however, a brief digest of points made so far.

First, the Thomistic understanding of hierarchy affirms that intrinsically valuable human goods are heterogeneous in character. These goods are thus not commensurable, such that the many goods we perceive are merely instantiations of, or instrumentally useful for, the absolutely ultimate end, the one and only intrinsically valuable human good and the single measure of goodness. This kind of commensuration has been roundly repudiated by contemporary writers and rightly so. Second, the incommensurability or heterogeneity of the human good, its resistance to any form of commensuration, does not preclude its being ordered according to priority and posteriority. All of being is ordered to substance, an order which is manifest in the moral sphere in the way that the different substantial human goods are ordered to that which is the substance of the human good in the most perfect sense: the absolutely ultimate end.

Third, within this hierarchical ordering of goods there are many different goods which are ultimate or final, though in a qualified sense. But the fact that they are ultimate in any sense means that they are desirable for their own sake. We don't just desire them because they help us achieve the absolutely ultimate end (though they do that, too); we don't wholly reduce them to instrumental goods simply because they have an instrumental aspect to them.

The Hierarchy of Happinesses

It is a commonplace of much contemporary Aristotelian scholarship and even some Thomistic that the for-the-sake-of relationship that I have been arguing is essential to an understanding of the human good cannot be applied to the ultimate end itself. In his book on Aquinas, for example, John Finnis speaks of beatitudo as a basic good. "But this turns out to be," he writes, "not so much an item to be added to the list of basic human goods, as rather a kind of synthesis of them: [namely] satisfaction of all intelligent desires and participation in all the basic human goods.and thus a fulfillment which is complete and intergral."{20} Elsewhere Finnis describes imperfect beatitude, at least, as "the good of complete reasonableness in one's willing of human goods."{21}

Roughly, then, we might summarize this view of the ultimate end as the activity of pursuing the basic goods in as unified a manner as possible, according to the hierarchy we have constructed for ourselves, without in any way disrespecting the intrinsic value of any of the goods by manipulating it as a mere means. To act in this way is to act both from reason and from virtue -- and, apparently, "for the sake of" the ultimate end of imperfect happiness.

But it is important not to confuse the "for-the-sake-of" relationship with this inclusive relationship described by Finnis. By an inclusive relationship I mean a formal relationship of part to whole, as when we say putting is a part of golf, or the pursuit of play is a part of imperfect happiness. To say that putting is "for the sake of" golf makes hash out of the phrase "for the sake of," if all that is meant is that putting is a constituent part of golf. True enough, putting is a part of golf if we are considering the constituent items and activities that go to make up a game of golf, like driving, chipping and cursing. This kind of consideration is formal in character. But within this formal consideration putting is in no meaningful sense "for the sake of" golf; for putting in this sense simply is golf. So, the only meaningful use of the phrase "for the sake of" is when the phrase is meant to say that one thing helps bring some other thing extrinsic to it into being and is regulated by it. Therefore, my two-putt on 17 is good insofar as it serves as the end of my desire to play golf on a Saturday morning; "putting" in this sense is the goal of an entire train of instrumental activities "for the sake of" landing me on the golf course. My two-putt on 17 is also good in an instrumental sense, insofar as it brings about, is "for the sake of," my continued play on 18 and my over-arching goal of finishing my round in the lowest possible number of strokes. "Putting" has an instrumental relationship to the good of "golf," if by "golf" we mean the completion of my round, not the game formally considered.{22} My point here is not to deny that it's meaningful to say, "Happiness for me is my family, my friends, my work, my recreational activities, my devotion to God, etc." We speak in this way all the time. My point, rather, is that this way of speaking does not refer to the primary sense of happiness, the sense which establishes the per se order of multiple goods according to prior and posterior -- and this sense, of course, is founded upon God. Without this sense of happiness, no other sense of happiness (including the reference to a set of constituents thereof), and more importantly, no direction for choice, is possible.

I have no intention of canvassing here Aquinas's arguments for the very existence of an ultimate end, for why the ultimate end must be one and not many, and the dialectical arguments he uses to manifest the nature of both imperfect and perfect happiness. Instead, I simply want to highlight some features of the arguments Aquinas develops in pursuing these questions.

First, when Aquinas argues in article 4, question 1 of the Prima secundae that there must be an ultimate end of human life, he stakes his claim on the fact that for human action even to get up and running, whether in the order of intention or of execution, there must be a per se order of ends culminating in an absolutely first, absolutely ultimate, end. And when in the next article Aquinas proves that the absolutely ultimate end must be single, he tells us that the hierarchy of "for-the-sake-of" relationships must culminate in a final end that is never for the sake of some other good, and thus serves to regulate every other good in the hierarchy which is, in some respect (though not in every respect) instrumental to it. In the sed contra of article 5 Aquinas glosses Matthew 6:24 ("No man can serve two masters") as a way of saying that no one can pursue two final ends not ordered to one another; that is, not situated within a network of "for-the-sake-of" relationships. Apropos of this latter text Germain Grisez has argued that "Of course, in choosing, one seeks a good loved for itself. In this sense, one always acts for an ultimate end -- that is, an end not pursued as a means to some ulterior end. But an ultimate end in this sense need not be the complete good of the human person, as Thomas assumed when he tried to prove that one's will cannot be directed simultaneously to two or more ultimate ends" [Grisez's citation is then to ST I-II, q. 1, a. 5].{23} But it is not the case that Aquinas didn't understand that there could be many final ends, each one imperfectly fulfilling of the human person. It is rather that he understood these less-than-absolutely final ends as existing in a duplex ordo to one another and to an absolutely ultimate end, in the absence of which no pursuit of any final end would ever occur. For what gets human action up and going is the pursuit of complete fulfillment of desire.

Human happiness is thus best defined as a unity in multiplicity, and this in more than one sense. As governed by the precepts of the natural law, our pursuit of happiness is always for something that is, in some sense, common. But as there are degrees of finality or perfection in goods in general, so there will be degrees of perfection in common good. There is only one common good that most perfectly satisfies the criteria of human happiness, and that is, strictly speaking, God himself. Accordingly, in the most perfect sense happiness is union with God in the next life; far less perfectly, it is contemplation of God in this life. But because we are not angels but embodied souls, this latter, mundane happiness must include the exercise of the moral and artistic virtues as ordered to the happiness of contemplation. This brings out the fact that a more perfect sense of happiness always subsumes that below it: all the happiness we seek in natural goods is taken up into and perfected in God{24}; while all the happiness we seek in the practice of moral and artistic virtue is taken up into and perfected in the life of contemplation.

Observations such as these customarily elicit two objections. The first objection I shall call, borrowing a phrase from Russell Pannier, the "personal destinies" objection. Does this objective hierarchy of goods leave any place for personal predilection and native talent in determining one's happiness?

What the good obliges me to do is structure my commitments according to the hierarchical framework of goods, rules and virtues. However, the basic framework can be instantiated, can be determined by the judgment of prudence, in myriad ways. I may possess neither the desire, talent nor opportunity to be a statesman, but justice will still be a good that I am bound to pursue. I may have neither the talent nor opportunity to study philosophy in a rigorous way, but I can still make contemplative activity the highest and best good that I pursue, perhaps by reflection on works of art, or conversations with friends, or by prayer. This solution to the problem does not deny -- indeed, it does everything to affirm -- a hierarchy of offices and duties that is not the production of individual choice. The wider a common good a particular office or duty looks after, the more divine-like and honorable it is.{25} Yet again, this does not mean that the lives of those who occupy lower offices are diminished. They are perfect in their own order, and in their perfection make a necessary contribution to the common good of the whole.

But what then of the related, "domination" objection? If contemplation and religious observance are the best goods, why shouldn't I spend all my time with them? To answer this we need to recall that higher goods in a hierarchy do not undermine the intrinsic goodness of the goods subordinate to them. My obligation to honor my parents, for instance, binds me to the goods of family life in a way that is constitutive of my happiness. My other obligation to honor God in the practice of the virtue of religion is not a rival to this obligation, even while it remains the more important obligation. The natural law in no way requires that I pursue religious acts to the exclusion of all other obligations. The natural law only demands that the religious obligation is given foremost respect in the tailoring of the hierarchy to my individual circumstances. In fact, it would be contrary to the proper understanding of my religious obligation if I did not understand the way in which it depends upon my lower obligations. The honoring of parents and the enjoyment of the goods of family life not only have their own requirements, but the intellectual and moral education one receives in participating in these goods is required if the religious good is to be fully achieved.

Particular Obligations

None of this discussion has yet identified any particular obligations. So in this final section I want to consider a bit more fully how the hierarchy I have been discussing issues in particular obligations, and in particular, I want to look at three obligations that arise at different levels of specificity in the order of practical reason.

Begin with the example of the golfer torn between his golf game and saving a drowning child. George's contention is that there is a moral rule that clearly says what one must do in such circumstances, and that rule is the Golden Rule. But what is the nature of this rule? What justifies it? George invokes it without explanation in the essay concerned with this golfing example. Finnis says more, at least about the rule itself: "The principle of love of neighbor-as-self," he argues, "and its specification in the Golden Rule, immediately capture one element in . . . integral directiveness: the basic goods are goods for any human being, and I must have a reason for preferring their instantiation in my own or my friends' existence." But why doesn't our golfer have a reason to prefer his life, and his golf, to that of his neighbor? The reason can only be explained in terms of hierarchy, and in the following way.

It is not at all contrary to my pursuit of a hierarchy of common goods to pursue some goods that are not in any immediate sense shared with the other members of my community. To go back again to the first lectio of the Commentary on the Ethics, Aquinas says that the whole which the polity or family constitutes has only a unity of order (as opposed to an absolute unity of composition, conjunction or continuity).{26} A family or polity is not a substance in the strictest sense, where every operation of a part is necessarily an operation of the whole. As a unity only of ordered parts, existing in "for-the-sake-of" relationships, one member of the order may have an operation that is not the operation of the whole, just as a soldier may have an activity (a furlough) that does not belong to the army as a whole. Nonetheless, the activities of the individual members of a unity of order are necessarily subordinated to the good of the whole. The soldier's furlough is ultimately for the sake of the army's victory over the enemy, insofar as the furlough refreshes the soldier for his duties. Indeed, the good of the whole demands that at times an individual seek a personal good both for his sake and for the sake of the whole. In the same way, our golfer's leisurely round is both for his sake and for the sake of his common pursuits with other members of his family and of his polity. But in those circumstances where a necessary feature of the common good is in jeopardy, as when a fellow citizen's life is in danger, a clear principle is invoked: the inferior, individual good must be abandoned for the sake of the higher, common good.

This golfing example deals with a moral principle at practical reason's highest level of generality, the principle that obliges us to love others as we love ourselves based upon the hierarchy of individual to common goods.{27} Let's look now at another obligation, this one from practical reason's middle range. In the argument against polygamy that we find in the Supplement to the Summa, Aquinas sets down this principle: "Whatever renders an action improprotionate to the end which nature intends by a certain work, is said to be contrary to the natural law."{28} Aquinas then distinguishes that an action may be improportionate either to the principal or secondary end of an action. First, on account of something which wholly hinders the end, as a very great deficiency in eating hinders both the health of the body (the primary end of eating), and the ability to conduct our business (the secondary end of eating). Secondly, an act may be improportionate to either the primary or secondary end of an action by making its attainment difficult, or less satisfactory. I take this distinction to be a restatement of what we read at Prima secundae q. 99, article 1: our obligations under natural law bind us to whatever is absolutely necessary or expedient for the sake of the ends to which nature directs us. Any action out of line with this necessity or expediency is improportionate to the end, and therefore illicit. According to Aquinas, the principal end of marriage is procreation, and its secondary end is, in a word, the bonum fides shared between the spouses. Marriage also has a third, sacramental end, namely, the signification of the union between Christ and the Church.

Now when it comes to the question of polygamy, a plurality of wives in no way hinders or makes inexpedient the primary end of marriage. A man can just as well beget children from one wife or many. But it's a different story when it comes to the secondary end of marriage, the bonum fides. A plurality of wives, Aquinas says, while not wholly hindering the shared life of the spouses, hinders it greatly, as a man cannot easily satisfy the requests of several wives, and because the sharing of many in one office causes strife. And when it comes to the sacramental end of marriage, polygamy destroys the signification altogether. So, polygamy as a means for the sake of the primary end of marriage is not against the natural law, and clearly this conclusion arises out of the way in which the means of polygamy is a perfectly suitable instrument for the begetting of children. But polygamy is against the secondary principles of the natural law because it greatly hinders the secondary end of marriage. For these reasons Aquinas concludes that polygamy both is and is not contrary to the natural law.

Interestingly, however, the case of polygamy also helps us round out the argument by providing an example of where obligation arises out of the deliberations of prudence. In the succeeding article in this question of the Supplement, Aquinas goes on to discuss whether polygamy was ever lawful. Polygamy, he says, does not trespass against the first precepts of the natural law, because it does not hinder the primary end of marriage. But again, it does trespass against the secondary precepts, precepts that Aquinas says hold not always but in the majority of cases, because it greatly hinders the secondary end of marriage. Aquinas affirms that this secondary precept is framed by God and even written on the human heart, but that it was dispensed by God through an inward inspiration to the holy patriarchs at a time when it was expedient to dispense with the secondary precept. Why? Because the end of the primary precept, the begetting of children necessary for building up the kingdom of God, was, at that time, of over-riding necessity. So here we see how the decisions of prudence, in this case divine prudence, depend upon the discernment of hierarchy: the primary end of marriage regulating the pursuit of the secondary end of marriage in circumstances where the primary end is in jeopardy. The understanding, moreover, of the primary good's being in jeopardy depends in turn upon seeing how the primary end of marriage is for the sake of, and thus regulated by, religious observance.


In this essay I have depended upon some central texts of Aquinas on obligation in order to show how the concept of a hierarchy of goods gives real direction to human choice. According to Aquinas, obligation arises first of all out of the recognition of an intrinsically valuable good to which we are naturally ordered, and secondly out of the recognition of means that either are necessary or expedient to the attainment of that good. Hence obligation depends upon recognition of hierarchy among goods.

Three features of this hierarchy have been particularly important for this discussion. First, that hierarchy fundamentally consists in asymmetrical causal relationships, in which a lower good is for the sake of a higher good in an efficient causal sense, but which also is regulated by that higher good in regard to the whether and how of its pursuit. Second, that this hierarchy of for-the-sake-of relationships culminates in an absolutely ultimate end that is not merely ultimate in an inclusivist sense, but which gives overriding direction to human choice by being the best good that we desire. Third, and perhaps most important for the debate with George, is the fact that in this hierarchy there is a wide range of ultimate ends that manifest intrinsic value while still being in certain respects for the sake of intrinsically more valuable ends, and most of all for the absolutely ultimate end. The instrumental aspect of these goods is an integral part of their nature as goods, even though it is not the only part of their nature. Still, when it comes to choices between these less-than-absolutely ultimate ends, between what George calls basic goods, the order brought into being by their instrumental relations is indeed what makes possible non-arbitrary direction for choice.


{1} Robert P. George, In Defense of the Natural Law (Oxford University Press, 1999), Chapter 2, note 125, pp. 81-82.

{2} Ibid., p. 93.

{3} Ibid., pp. 93-94.

{4} Ibid., p. 98.

{5} "Invenitur autem duplex ordo in rebus. Unus quidem partium alicujus totius seu alicujus multitudinis adinvicem, sicut partes domus adinvicem ordinantur. Alius est ordo rerum in finem. Et hic ordo est principalior, quam primus. Nam, ut Philosophus dicit in undecimo Metaphysicorum, ordo partium exercitus adinvicem, est propter ordinem totius exercitus ad ducem." In I Ethicorum, lectio 1, no. 1.

{6} See especially ST I-II q. 99, a. 1, and II-II q. 44, a. 1.

{7} The incommensurability between what he calls "agent-relative" and "agent-neutral" reasons for action is defended by Mark C. Murphy in Natural Law and Practical Rationality (Cambridge University Press, 2001), esp. p. 186.

{8} ST II-II q. 26, a. 1: "prius et posterius dicitur secundum relationem ad aliquod principium. Ordo autem includit in se aliquem modus prius et posterius. Unde oportet quod ubicumque est aliquod principium, sit etiam aliquis ordo.."

{9} A problem discussed by Hittinger, A Critique of the New Natural Law Theory, Chapter 2, especially pp. 73ff.; and Benedict M. Ashley, O.P., "What is the End of the Human Person? The Vision of God and Integral Human Fulfillment," in Luke Gormally, ed., Moral Truth and Moral Tradition: Essays in Honor of Peter Geach and Elizabeth Anscombe (Dublin: Four Courts Press, 1994), pp. 68-96.

{10} Murphy, Natural Law and Practical Rationality, p. 192.

{11} EN I.6 1096a17-23. Cf. In I Ethicorum, lectio 6, nn. 79-80. On this point I have learned from Kevin L. Flannery, S.J., Acts Amid Precepts (Washington D.C.: The Catholic University of America Press, 2001), Chapter 4. It is interesting to relate Aristotle's argument to Aquinas's analysis of the goodness of the human act. The species, or substance, of a human act, according to Aquinas, is a form/matter composite. This composite is considered formally in terms of the end (the object of the interior act of the will), but materially in terms of the object of the exterior act. The circumstances of the act accrue to the substance of the act as accidents of it. Thus in the human act there is an order of priority and posteriority, the substance of the act being prior. See ST I-II, q. 18, a. 6 and q. 7, a. 3.

{12} Murphy, Natural Law and Practical Rationality, p. 192. Here Murphy also references John Finnis, Natural Law and Natural Rights , pp. 100-105. See also Russell Pannier, "Finnis and the Commensurability of Goods," The New Scholasticism 61 (1987): 440-461, esp. 443. Unlike Pannier, and for the reasons already given, I do not think it accurate to refer to a hierarchical ordering of goods as a commensuration of those goods, except perhaps in a very loose sense.

{13} Ibid.

{14} In I Ethicorum, lectio 1, no. 1.

{15} Henry S. Richardson has a nice discussion of the way Aristotle uses counterfactuals in distinguishing final ends from instrumental ends, as well as from final ends that are also instrumental ends. See his Practical Reasoning About Final Ends (Cambridge University Press, 1997), pp. 53-57.

{16} Richard Kraut, Aristotle on the Human Good (Princeton University Press, 1989), Chapter 4.

{17} In I Ethicorum, lectio 1, no. 27.

{18} This kind of argument is developed in Alasdair MacIntyre's defense of a teleological, as opposed to a functionalist, account of virtue in his "S(phrosun(: How a Virtue Can Become Socially Disruptive," in Midwest Studies in Philosophy XIII (1988): pp. 1-11.

{19} George, In Defense of the Natural Law, p. 96. See also Russell Hittinger, "After MacIntyre: Natural Law Theory, Virtue Ethics, and Eudaimonia," International Philosophical Quarterly (December 1989): pp. 449-461.

{20} John Finnis, Aquinas: Moral, Political and Legal Theory (Oxford University Press, 1998), pp. 85-86.

{21} Ibid., p. 108.

{22} These remarks can also be taken as a partial rebuttal of J. L. Ackrill's inclusivist understanding of Aristotle's view of the good. See Ackrill's "Aristotle on Eudaimonia," reprinted in Amélie Rorty, ed., Essays on Aristotle's Ethics (Los Angeles, CA: University of California Press, 1980).

{23} Germain Grisez, The Way of the Lord Jesus, vol. I, Christian Moral Principles (Chicago: Franciscan Herald Press, 1983), pp. 809-10, quoted and discussed in Ashley, "What is the End of the Human Person? The Vision of God and Integral Human Fulfillment," pp. 68-69, especially.

{24} It is interesting to note in this regard that for Aquinas even the moral virtues endure after this life, albeit in their formal, rather than material nature. See ST I-II, q. 67, a. 1. Something analogous holds with the intellectual virtues, including, I assume, the virtues of art (ST I-II, q. 67, a. 2).

{25} Cf. In I Ethicorum, lectio 2, no. 30.

{26} Ibid., lectio 1, no. 5.

{27} Key Thomistic texts concerning this principle are ST I-II q. 99, a. 1; q. 100, a. 3, ad 1, and a. 11.

{28} ST, Supplement, q. 65, a. 1.