Finally, the Aquinas-Porter argument can be revised in a way that is not liable to the objections to the first step and the second. Suppose we say that the agent’s encompassing good is not some more general good, like the good of the polis or of the natural world or of God’s friends, that necessarily includes her individual good, but the divine itself, which is by its own nature self-transcending. For Scotus, the end for human beings is to enter into the love that the persons of the Trinity have for each other, or to become co-lovers. This means that the highest activity is one of will, prepared by intellect, since Scotus accepts the principle that nothing is willed except what is previously cognized. This view should be distinguished from a view that happiness involves both the intellect and the will, but the activity of the will (the loving) is consequent on the highest activity, which is the beatific vision in the intellect. On Scotus’s view we can say that, of faith, hope, and love, the greatest is love both here and in heaven. On the other view, love may be the greatest of the three down here on earth, but in heaven the greatest will be a state of the intellect. The proposal we are now considering is that the agent identifies her happiness as entering into a kind of loving that is itself self-transcending. Such divine self-transcendence can be seen within the Trinity.

The proposal we are considering would appear to overcome the difficulties with the two steps of the Porter-Aquinas argument. First, it doesn’t beg the question about motivation. It allows that we have both a self-perfecting and a self-transcending love. But it holds that the second comes out of the first, because we identify in perfecting ourselves with a being that is itself self-transcending. Second, it does not hold that there is a necessary harmony between the self-indexed interests of an agent and the wider groups in which she is included.

The proposal has been stated in theistic terms, but can also be found in non-theistic terms. It might seem that if you identify the best (the most perfect) state of yourself as loving, then there will no longer be a tension between perfecting yourself and spending yourself for others. But there’s a difficulty here, which discloses itself in the following dilemma: either this is a single-source theory (deriving all motivation from my own happiness and perfection) or it is not. We have interests that do not reduce to virtue, or to conforming our lives to the Categorical Imperative for its own sake, and it’s completely appropriate for beings like us to have these interests. The point was made earlier that Kant should have allowed that self-indexed motivation includes more than the satisfaction of sensuous inclination. Now a single-source theory can to some extent accommodate such motivation. Aristotle, for example, can insist that self-love of the right kind is consistent with various forms of self-sacrifice. But he also insists that these are forms of self-love. His picture of motivation is that, if the agent were to ask herself “Why am I doing this?” the fundamental answer would be “because I am assigning myself the best thing.” Scotus and Kant would say that this answer is unacceptably self-regarding. But there is a dilemma here. Some followers of Aristotle disagree with his point about self-love and say the virtuous person should be called “good-loving.” This is unobjectionable, but now we no longer have a single-source theory. In Scotus’s terms, God will be loved both for God’s own sake and for the sake of the union. So either we stick with a single source theory that seems objectionably self-regarding, or we allow that self-indexed motivation properly remains, and then we have a double-source theory again, like that of Kant and Scotus.

The point is that the self-indexing of some goods needs to remain. An attempt to get rid of all such goods is Maimonides’ notion we’re absorbed into God. But surely there’s a sense in which we lose ourselves on that view; we lose, in Scotus’s term, our haecceity. One way to put this is that the fourth defense of eudaemonism paradoxically ends up compromising the aspiration to happiness.

So there is a dilemma for this kind of “agent-perfective” account of eudaemonism. It is the best form of eudaemonism; one free from many of the objections raised in this chapter. But it still faces the present dilemma. Suppose we think that an agent should be motivated by the desire to perfect herself, and suppose that being perfected is becoming the kind of person who is not always motivated by self-indexed goods. Do we now have a form of eudaemonism that is not “unacceptably self-regarding”? The problem is that we need to know whether this is a single-source account of motivation. If it is, and this account does not retain motivation toward goods that are self-indexed and necessarily so (such as the particular way of loving God that is unique to an individual), goods that could be (counterfactually) in tension with God’s own good, then the account is, we might say, “unacceptably self-neglecting.” But if it keeps these goods, then it is no longer a single-source account. By the definition at the beginning of this chapter, this means it’s no longer a form of eudaemonism. But what we need in our substantive theory is an account that gives us both kinds of goods.