Love and Obedience

Alexander R. Pruss

Department of Philosophy

Georgetown University


To be presented at the Eastern Regional Meeting of the Society of Christian Philosophers, Columbia, SC, April 20, 2007


1. Introduction

            As Mark Murphy has recently shown, standard justifications of universal divine authority are insufficient.[1] By “divine authority” I shall mean the doctrine that obedience is morally owed to God by all. God would not give us a command that we did not have a reason to act in accordance with, Murphy argues, but it does not follow that we would be obliged, much less morally obliged, to have the fact of God’s having commanded the action be among our practical reasons for the action. We might, for instance, think that what God has commanded us is prudentially the best policy for us or that it is the best possible evidence for what is on independent grounds morally obligatory.

            Particularly illuminating in this connection is Murphy’s refutation of arguments in favor of divine authority which have the form:

(1)  Everyone ought to stand in relation R to God.

(2)  Everyone who stands in relation R to God is obliged to obey God.

(3)  Therefore, everyone is obliged to obey God.

For instance, the relation R may be loving or being grateful or begging forgiveness of. We can see that in general such arguments are formally fallacious, since they have the following logical form:

(4)  x ought to do A.

(5)  If x does A, then x ought to do B.

(6)  Therefore, x ought to do B.

We can see the fallacy by example: Under some circumstances a surgeon ought to cut away a tumor on the patient’s skin, and if she removes the tumor, she ought to suture the patient. But supposing she does not actually perform any surgery, her failure to suture the patient is not an additional fault.

            However in cases in which doing B is partially constitutive of doing A, the obligation to do A does seem to entail an obligation to do B. Moreover, I shall argue that love for God is partially constituted by obedience to God, and hence the obligation to love God entails an obligation to obey God.

            The version of the divine authority thesis that I will argue for is that God’s will, or more precisely an aspect of God’s antecedent will, is owed obedience.[2] In this way, obedience to God may be different from obedience to civil authority. For there we obey the command rather than the will of the legislator.[3] But practicing theists do think and talk about “discerning the will of God”.

Now one tempting but mistaken rendering of the thesis of divine authority is that God knows the best thing for us to do, and hence we ought to obey God. But this only gives God epistemic and not practical authority, and gives us evidence of further reasons to act, but does not give us a direct reason to act out of obedience. On the account I shall defend, divine commands also have this kind merely evidential authority. However, what a divine command is evidence for is a claim about God’s will, and God’s will has practical authority—it is to be obeyed.

Much of this paper will be devoted to developing an account on which love for God is partially constituted by obedience to God. I shall do this by examining the way that reasons for to those whom we love as “other selves” constitute reasons for us so that what the beloved self-legislates for herself is equally authoritative for the lover. The account of reasons for the beloved is inspired by Aquinas’ description of what he calls the “formal union” of love, whereas self-legislation constitutes a Kantian element (albeit with Thomistic antecedents). It will follow that when one loves someone who is superior in relevant respects, including in respect of having already duly considered the inferior’s reasons, the inferior owes obedience. This will also yield a partial defense of Aquinas’ otherwise puzzling view that lower angels are obliged to obey higher ones, and we are obliged to obey even the lowest unfallen angels. And it will turn out that everyone is required to obey the will of God.

One price to pay for an account that does all of these things is that it is a mere sketch, dogmatic in portions. The dogmatic sections, however, can be epistemically justified by the fact that the account solves the puzzle about why we are obliged to obey God better than alternatives do.

This paper is part of a general project to ground all ethics in the duty to love.

2. Constitutive parts of obligations

            If I am obligated to do A, I am also obligated to do what it takes to do A when this is morally possible. I cannot excuse my failure to pay what I owe you because I could not pay my debt without having your address whereas I had taken good care to lose that address. “What it takes to do A” includes both means and constituents. Thus, holding on to your address is a part of what it takes to pay my debt, since it is a means to the payment. But holding on to your address is not a constitutive part of paying the debt. On the other hand, giving you legal title to a sum of money is a constitutive part of paying the debt, and another constitutive part is rendering the sum of money accessible to you.

In general, from my being obligated to do A, and from A being such that, necessarily, B is partially constitutive of A, I can conclude that I am obligated to do B. We are, after all, obligated to do what it takes to fulfill our duties. All I mean by “obligated” is that there is there is a conclusive moral reason. It is clear that if there is a conclusive moral reason for engaging in an activity, this is equally a conclusive moral reason for performing its constituent parts.

3. Love and authority

            According to Aquinas, a constitutive part of loving someone is what he calls a formal union in affection (Summa Theologiae I-II, 28, 1). Moreover, Aquinas says that an effect of love is a mutual indwelling of intellect and will (I-II, 28, 2). The person I love comes to be present, as it were, in my intellect and my will, in that I think about the person and will his good. At the same time, I come to be present in my beloved in intellect and will, in that I seek to “[penetrate] into his heart, as it were”, see the reasons that he has for action as reasons for myself, and take goods and bads happening to him to be my goods and bads. I shall take this mutual indwelling to be partially constitutive of love.[4]

            If I love someone, then the love impels me towards what Aquinas calls “real union”, for instance spending time with the beloved. It may even be that the love gives me a duty to engage in real union. However, the real union is not a constituent of love, and hence the obligation of real union is conditional on the actual presence of love. But on my reading, mutual indwelling is a part of love, and hence if I am obliged to love someone, I am thereby obliged to engage in the mutual indwelling.

            I take it to be a basic fact of ethics that we are all obliged to love everyone, though my main argument only needs the weaker claim that we are obliged to love God. That we have a duty to love God can be argued for on the basis of his supreme lovability, for instance. It is plainly required to love someone: a life without any love is clearly lacking something that ought to be there. But any reason to love someone (whether understood in the sense of a reason to love someone-or-other or of a reason to love a particular someone) is matched by an even better reason to love God, since God is more lovable than anything other than God. We might also argue for the duty to love God out of gratitude for creating us, say. Or we can just take the duty to love God as the most basic truth of ethics.

I shall, further, only assume that I am obliged to love God at least as much as I love myself. The surprise will be that this condition is sufficient to generate divine authority. Of course it is plausible that I am obliged to love God much more than I love myself, but I will not need the stronger claim.

            It is morally required, thus, that I engage in mutual indwelling with God and love him at least as myself. I will now argue that mutual indwelling constitutively involves a particular way of obeying the beloved’s own reasons.

            Recall that in the indwelling, I must will the beloved’s good “from the inside”, as it were. The beloved is “another self”, and his goods are my goods, his reasons are my reasons. Suppose now that I love my beloved at least as myself. A constitutive part of that love is, roughly speaking, treating my beloved’s goods as at least as important to me as my own goods, and weighing my beloved’s reasons at least on par with my own reasons in my deliberation.

Here, however, there is a crucial point to be observed.[5] Were I to say that when I love someone, I make his reasons become reasons for me, my overall argument would fall prey to the fallacy identified at the beginning. For from the obligation to love God it might follow that I would have an obligation to make his reasons become reasons for me. But it would not follow that his reasons would have any authority for me if I did not actually love God and did not make his reasons become reasons for me. However, my point is not that a constitutive part of loving someone at least as myself is to make someone else’s reasons become my reasons. Rather, it is a constitutive part of love to act on the other’s reasons as appropriate and to weigh them at least on par with mine in deliberation. If I have a duty to love the other person, then the other’s reasons are reasons for me whether or not I actually act on them or weigh them in deliberation, and they are reasons in virtue of the fact that my acting on them and weighing them is a constitutive part of the fulfillment of my duty to love that person.

I will only use the word “friend” in cases where I have good reason to love the person. I actually take it that I have good reason to love everyone, but I do not need that claim: in the end I will apply my account to God.

Now we need to tread more carefully in the case of essentially agent-relative reasons, reasons such that the value of promoting the end in question depends who promotes it. Thus, if the man I love has a reason to kiss his wife, this does not imply that I have a reason to kiss anybody, since the value to my friend of his wife’s being kissed is entirely dependent on who does the kissing.

On the other hand, imagine that my friend’s wife is in prison, so that giving him an opportunity to kiss her takes some arranging. The value to my friend of his having an opportunity to kiss his wife is not in the same way entirely dependent on who provides this opportunity. If the value of the opportunity is no way dependent on who arranges the opportunity, then if I have reason to love my friend at least as myself, the objective weight of the reason for me to arrange that opportunity is at least as great as it is for him.

In general, the value of some goal that my friend has will to some extent depend on who achieves the goal. There will be goals where it is preferable to him that the goal not be achieved by him but by me (e.g., if my friend wants reassurance), and there can be goals where it is preferable that they be achieved by a complete stranger. In having a reason to love my friend at least as myself, I have a reason to promote a goal and the objective weightiness of the reason is at least proportional to the value that my promoting the goal in question has to the friend.

When I talk of the value that something has to someone, I am speaking objectively of how much he would benefit from it. My reason to act is dependent on the actual value of the goal to my friend, not on my friend’s perception of its value to him. We often use the phrase “G has value v for x” to affirm that x (consciously or not) holds G to have value v. But here I am specifically using it in a sense that does not in general entail any beliefs by x about the value of G. Friends can learn from and correct friends.[6] If my friend wrongly believes that something is good for him, then just as he has no reason to promote that good, but merely thinks he does, so, too, I do not have a reason to promote that good.

Observe that loving my friend as I love myself will sometimes constitutively involve my abandoning my own goals in order to fulfill his. Reasons that I have in my own right are often unequal between each other, and I may need to act on a stronger reason rather than a weaker. It can, likewise, be the case that my friend’s reason for promoting a goal is actually stronger than my own reason for promoting some incompatible goal. In such a case, if my friend’s reason does not depend on his being the one to promote that goal, then I should see the reason for promoting his goal as stronger than the reason that I have in my own right for promoting the incompatible goal.

This implies, surprisingly, that there could be a case of a friend whom I love as myself but all of whose goals override my own, even when I only love my friend as myself, if these goals have a greater value to my friend than my own goals have to me.

4. Autonomy

But if that were the whole story, then the need to see things from the beloved’s point of view would be largely instrumental to figuring out what is, in fact, objectively valuable for the beloved to achieve. The story as far as it goes is correct, but incomplete. For there may be certain goals that are valuable to my friend in part because he has set them for himself, and not just because they are antecedently mind-independently valuable to him. Here seeing things from the beloved’s point of view is crucial.

Now my friend may have Kantian autonomy, an authority over himself that lets him set laws for himself, creating new reasons for his own actions. Suppose, then, that such a friend commits himself to vegetarianism, believing that vegetarianism is not morally required but nonetheless a good way of affirming the value of all animal life. Insofar as through this my friend has given himself a new reason to make it possible for himself to live a vegetarian lifestyle, this in turn gives me a new reason to make it possible for him to live a vegetarian lifestyle. I have as much reason to prepare a non-meat dish for him when he visits me as I would to prepare one for myself were I such a vegetarian.

If my friend has Kantian autonomy in some area of life, then by willing certain ideals in that area, my friend brings it about that these ideals objectively have an additional value for him. This creates what one might call autonomistic reasons for him to promote these ideals. And this in turn gives me a reason to promote these ideals.

At the same time, an attempt to create autonomistic reasons might fail. The ideals in question might simply be bad, and the apparent exercise of autonomy might simply be invalid, just as a command to do something immoral is invalid and has no authority at all. Or one might be wrong in thinking that one has autonomy in this area of life. No reasons arise then from the attempted exercise of autonomy.

We can tell a story about what would give rise to the relevant kind of autonomy, a story following the Thomistic account of positive law as a determination of natural law. Unless something like utilitarianism holds, morality by itself is insufficient to determine our lives. For instance, we can be faced with incommensurable goods: a life of astronomy and a life of sculpting are both good, with neither being better than the other, and we may be capable of both. Moreover, it may be necessary for leading a life with the right kind of integrity and commitment to chosen ideals that one be able to set one’s will on an ideal such that one gains additional reason to pursue that ideal.

Furthermore, the additional reason to pursue the chosen form of life may be exclusionary. Thus, if you legitimately commit yourself to a life of astronomy, you might make the first order goods to yourself of sculpting no longer be rationally relevant to your decision to persevere in astronomy, so you will no longer need to constantly weigh the pros and cons to yourself of astronomy over sculpting. Moreover, doing astronomy out of a commitment to it seems to have a value of its own: a life committed to astronomical research has a different kind of value from merely a life of astronomical research, even if one’s CV is the same. But the exclusionary reason against the advantages of sculpting then has weight for me if you are my friend. If your exercise of autonomy was valid, then the fact that sculpting would be valuable to you is excluded from my consideration of what way of life I should promote for you.

On Raz’s account of authority[7], a valid command gives two kinds of reasons. It gives a first-order reason for doing the commanded action because it was commanded, and a second-order reason for disregarding certain first-order reasons against doing that action. It is now becoming plausible that there could be cases where having a duty to love someone would give us a similar pair of reasons, though not in light of something’s being commanded, but in light of something’s being willed by the beloved, say as an ideal. When I pursue goals because they have autonomistic value to my beloved, I pursue these goals because my beloved wills them.

5. God

            Now we may be skeptical that people actually have Kantian autonomy in any area. For instance, I might think that I cannot or ought not bind my own will in such a way as to exclude some reasons from operating, because not knowing what the vicissitudes of life may bring, or what God may call upon me to do, I have to be open to constantly re-weighing my reasons. Or I might simply think that the kind of dignity involved in Kantian autonomy is not for the likes of you and me. Our autonomy needs to be limited lest it unduly interfere with the changes in our lives, with our ignorance of the situation, and so on. Moreover, our autonomy needs to be limited because we have our own place and role in the universe to fill, including duties of obedience to others.

Thus our attempt to create autonomistic reasons for ourselves may sometimes, often or always fail. It is plausible that we will always fail to create strong autonomistic reasons for ourselves, ones that should override all reasons to the contrary. However, it is also plausible that we can create weak autonomistic reasons for ourselves, i.e., ones that should override some but not all reasons to the contrary.

But any skepticism about autonomy should disappear in the case of God. While God needs to be true to his own nature, this fidelity is compatible with infinitely many options. Thus, God did not have to create a universe, but did. He could have created unicorns, but did not. God does not have to worry that his understanding of a situation falls short of the reality of it. He is not under anybody else’s authority. He is not circumscribed in any way.

Now, Christian theology traditionally distinguishes God’s antecedently willing something from God’s consequently willing it (e.g., Summa Theologiae I, 19, 6, resp. 1) God’s consequent will is always exactly fulfilled: something is willed consequently by God if and only if God actually concurs in it, and every positive fact about creation is one that God concurs in. However, what God antecedently wills are the values that he desires the universe to instantiate. Antecedent will is more akin to a desire or, better, a commitment to particular values. The standard example in the tradition is that God antecedently wills everyone to be saved (ibid.) It does not, however, follow that everyone is saved, for it may be that God also antecedently wills that only those who do not freely reject salvation, and hence does not consequently will that Jones be saved given that Jones in fact freely rejected salvation.

It is plausible that God’s antecedent will creates autonomistic reasons for God. Certain goals acquire a new value for God as a result of his antecedent will. Thus while mercy and strict justice are always intrinsically valuable, and probably incommensurable, God opted for a preponderance of mercy over strict justice in his dealings with human beings. He could legitimately have exercised his autonomy differently, opting for a preponderance of strict justice over mercy.

If I love God at least as myself, then what is valuable to God gives rise to at least a proportional reason for me. My setting my will in accordance with God’s goals, to the extent that these goals are valuable to God, is going to be a constitutive part of my love for God, since this holds generally for love of anybody.

But in the case where the beloved is God, this implies obedience to the beloved’s will. For suppose that God antecedently wills that I do something, in such a way as to bestow an autonomistic value on my doing it. This gives me a reason for the action precisely because of the value the action has to God, i.e., precisely because the God whom I love at least as myself has willed it in the way he has. My being deliberatively moved by this reason is a constitutive part of my loving God at least as myself. Thus, I have a duty to give deliberative weight to this reason given that I have a duty to love God.

Moreover, the reason in question may well override all my other reasons for action. I can show this in two ways. First of all, the autonomistic value of a goal derives from the dignity of the one who has that goal. Thus, the autonomistic value of a goal had by a mere human being whom I love merely as myself will not override autonomistic values of goals that I have set myself, since we all equally have human dignity. But if my beloved has a greater dignity, then the autonomistic value of his goals is, plausibly, proportionately greater. Hence if I loved someone with a greater dignity, say an angel, then the angel’s autonomistic goals might well override mine, at least when the angel put himself fully behind these goals.

The reason for this overriding would not be that I would necessarily be obliged to love the angel more than myself, but that the angel’s strongest autonomistic reasons have objectively greater weight than the strongest of my own autonomistic reasons, and hence when I weigh the angel’s reasons side-by-side with mine, without assigning them any added, the angel’s reasons may well override mine. To recall Mill, if I had within my own will the human desire to pursue philosophy and the piggish desire to wallow, the desire to pursue philosophy would have greater weight. If I love a friend as myself, then reasons deriving from his or my desire to philosophize should override reasons deriving from my or his desire to wallow.

Nonetheless, I may have non-autonomistic reasons, say ones coming from my duties to my family, that could override an angel’s autonomistic reasons. But in the case of God, things are different. God knows all my desires, plans and commitments, and in his all-loving deliberation, he gives due weight to them all. When God, then, sees an autonomistic reason coming from one of his goals as overriding some reason coming from a goal (autonomistic or not) of mine, he is right, being omniscient. His autonomistic reason, then, does have greater weight. Insofar as I love God at least as myself, in my deliberation I give my reason for pursuing God’s autonomistic goals a weight at least proportional to the value the goals have for God. If God sees this value as overriding the value that every goal I have in my own right, then I have a reason out of the duty to love to act to pursue that particular value which God, in his autonomy, has set for himself.

Now, not all of the reasons that God has antecedently willed need to be overriding in this way. Thus, God’s antecedent will that the laws of nature be followed is miraculously set aside by God from time to time due. It is possible for God’s antecedent will on some occasion to give a value to a goal, where that value to God is less than the importance to God of my avoiding suffering, and in that case love does not require me to pursue the goal when my suffering is the price. On the other hand, it is possible for God’s antecedent will to include goals that give rise to reasons that exclude some or even all other first order reasons.

If so, then a constitutive part of my love for God will be my obedience to a reason to pursue these goals, and hence may well include my obedience to second order reasons of the sort that are involved in Raz’s account of authority, if God has in fact willed such goals in such a way as to give rise to an exclusionary reason. Now, a command from God is a speech act that can be taken as evidence that he has in fact antecedently willed certain things to the exclusion of everything else. If so, then we are obliged to obey divine commands, not because they are commanded, but because they reveal to us the antecedent will of God.

Let me end, however, with a difficulty. God’s antecedent will may seem to cohere poorly with divine simplicity, since divine simplicity entails that God has no contingent intrinsic properties. For someone like me who accepts divine simplicity, there are only two ways out of this difficulty: (1) insist that God’s antecedent will is necessary and not contingent; or (2) hold that facts about God’s antecedent will are a combination of necessary facts about God and contingent facts about the worlds that God has created.[8] Given option (2), any divine commands expressed in creation will be a part of the reductive base, and this may yield the view that divine commands are to be obeyed also because they are commanded, since their being commanded may be partly constitutive of their being antecedently willed.

[1] An Essay on Divine Authority, Cornell University Press, 2002.

[2] Cf. Mark Murphy, “Divine Command, Divine Will, and Moral Obligation”, Faith and Philosophy 15 (1998), 3-27.

[3] Mark Murphy has noted in correspondence that even there, there is such a thing as obeying the “spirit of the law” and so the difference is smaller than it might seem.

[4] While Aquinas talks about these indwelling mutual indwelling relationships as “effects” of love, the term “effect” here is compatible with the mutual indwelling being partially constitutive of love, with the kind of causation here being formal rather than efficient causation. This is exactly how Aquinas explicitly understands “effect” in the context of formal union in affections (I-II, 28, 1). The mutual indwelling is discussed in a separate article from the discussion of formal union in affections, but it seems that at least a part of the mutual indwelling is that formal union.

[5] I am grateful to Mark Murphy for pointing out the danger here.

[6] Cf. Jennifer Whiting, “Trusting ‘First’ and ‘Second’ Selves: Aristotelian Reflections on Virginia Woolf and Annette Baier”, in: J. Jenkins, et al. (eds.), Persons and Passions: Essays in Honor of Annette Baier, Notre Dame University Press, 2005.

[7] Joseph Raz, The Authority of Law: Essays on Law and Morality, Clarendon Press, 1979, pp. 21-27.

[8] A similar claim concerning in effect God’s consequent will is made in the discussion of God’s free will by Jeffrey Brower, “Simplicity and Aseity”, forthcoming in The Oxford Handbook of Philosophical Theology.