The Origin of Moral Evil

Having seen in connection with the De Trinitate that for Augustine the image of the Trinity in man is not only a “heuristic tool for introspective theology” but also a moral challenge, we should first see what impedes us from living up to “the anthropological task” posed by that challenge.
This is the reason why, before moving on to the systematic discussion of the theological virtues of faith, hope, and charity in the Enchiridion, the virtues that would help us in the realization of our “anthropological task”, we have to return to a discussion of the origin of moral evil, as it emerges in the context of the third book of Augustine’s De Libero Arbitrio.
The main question then, is this: how does moral evil arise in human actions? Is it something inevitable? If so, doesn’t that eliminate our responsibility? If not, then it should be in our power to avoid it. But then, why don’t we?
In the first approach it might seem obvious that if the will is free and a human action is one that is determined by the free choice of the will, then it should be in our power to avoid evil in our actions, that is, to avoid sin.
But this seems to be in conflict with divine foreknowledge: if God knows that I will sin, then it cannot be otherwise, and what cannot be otherwise is inevitable, so, apparently, I will inevitably sin.
The simple solution to this problem is that God’s foreknowledge does not necessitate that I will sin: what is necessary is that God knows that I will sin, if and only if I will sin, but from this it does not follow that I will necessarily sin, for God only knows that I will sin if I will in fact sin, and if I will not sin then of course God does not know that I will sin, on the contrary, then he knows that I will not sin. But whether or not I will sin will be determined by whether I choose to do so, and that is again something that God necessarily knows if and only if I will in fact choose to do so.[1] So, divine foreknowledge of my actions does not take away the freedom of those actions, and so it does not take away my responsibility either. For I am always responsible for my actions precisely because and to the extent that they are my actions determined by my free choice.
But then, how come people choose to sin when they could avoid sinning? After all, isn’t it better not to sin? If so, why would anyone not choose what is better over what is worse? Aren’t people supposed be thinking before their choices and choose the best alternative? Perhaps there still is something that somehow makes the will “disoriented” and miss its goal, namely, choosing the best possible alternative.
In fact, this is precisely the line of reasoning behind two ancient answers to the problem of the origin of moral evil, both of which were explicitly rejected by Augustine. (a) The “Socratic” answer: what makes the will go wrong is our ignorance, some cognitive failure, on account of which we mistake the apparent good for the real good – so we rationally, but mistakenly, choose the lesser good, or something that is in itself evil; and (b) the “Manichean” answer: according to this conception, although the will in itself is good and striving after the good, sometimes it is simply overwhelmed by some evil, material principle, so it is some extrinsic compulsion (which may not be extrinsic to the man who has the will, but extrinsic to the will or the soul trying to fight this compulsion) that eventually makes a person do wrong.
Augustine rejects both of these answers, because these, instead of explaining the origin of moral evil, rather seem to provide excuses for the evil deeds of rational, free agents, whose responsibility is diminished precisely by these factors, namely, ignorance, on the part of our limited cognitive powers, and compulsion, on the part of our limited active powers.
So Augustine locates moral evil in the will’s free, deliberate act itself, when it does something wrong knowingly, deliberately, and in its full capacity to do otherwise. It is this deliberate, and intentional turning away from what is good, and pursuing what is bad, despite the best counsel of reason and the capability of not doing so that is the archetypal moral evil Augustine is after, first exemplified in the fall of angels, and then in the fall of man. ***
The conclusions of Augustine’s discussion can be summarized in the following points:
Since moral evil is the result of free agency, there is nothing that necessitates the occurrence of moral evil in creation.
However, the only free agent with absolute omniscience and benevolence is God, and only a free agent of this kind can have an absolutely immutable, unfailing good will.
Therefore, if there are to be free agents of limited perfection, with mutable wills, which can only be free if they can occasionally choose evil, then with the creation of such agents the possibility of moral evil necessarily arises.
But the realization of this possibility is entirely dependent on the autonomous abuse of their freedom: when these agents arrogantly assert their freedom against what is shown to them to be the best, when, like spoiled children, they rebel against Goodness itself, in their swollen pride. (So, moral evil simply arises as the “temper tantrums” of imperfect free agents.)
However, this arrogant pride cannot be broken by force, for that would eliminate the very point of having created these agents as free in the first place. What they really need is a bit of “growing up”, learning from their own faults as their lives unfold, until they realize that the only genuine way of exercising their freedom is freely wanting what Goodness demands, that is, by subjecting their will to divine will. This is how ethical rationalism and divine command ethics converge in Augustine’s thought, as being just different expressions of the same moral imperative. (And, by the way, this is how Kantian rationalism also finds the pure realization of the categorical imperative in the act of the holy will, in the voluntary act of the perfectly rational free agent.)
Therefore, the realization of this imperative is in fact possible through an individual agent’s finding and exercising the theological virtues throughout their lives, living according to the precepts that follow from them. These are the virtues discussed in detail in the Enchiridion.
However, since man is an essentially social animal, the individual human person’s history is inconceivable without the history of mankind, for each human person’s life is but a particular realization of the general human condition in its several historical stages.
Therefore, from this perspective, the whole of human history turns out to be nothing but the story of humanity’s “growing up”, realizing in its saints the human nature which freely, yet firmly, and unfailingly directs its will towards Goodness itself. But this realization is possible only through individual acts of free individuals, who individually may fail in countless ways. Indeed, even the saintliest may occasionally yield to temptation, whence they necessarily need God’s supernatural help for their salvation, namely, divine grace. But divine grace is effective in helping those who render themselves capable of receiving it.
Therefore, although “many are called, but few are chosen” (Matth. 20,16), for humanity in its fallen state is so blinded by its own arrogance that many do not even notice how they doom themselves to damnation, the history of humanity is still the history of salvation, which will be the grand vision of the City of God.

Cf. also the following questions from Aquinas.

[1] The response, therefore, simply points out that the argument trying to establish the incompatibility of free choice and divine foreknowledge is a non sequitur. For even if N(Kpp), this does not entail KpNp (de dicto/de re; sensus compositus/divisus; necessitas consequentiae vs. consequentis; De veritate, q. 24 a. 1 ad 13; Super Sent., lib. 1 d. 38 q. 1 a. 5 ad 3), except in the sense of KpNpp. The real challenge, however, is this (Super Sent., lib. 1 d. 38 q. 1 a. 5 arg. 4):N(Kp); N(Kpp)/Np. Note, however, that pN(Kp) is false; only this is true: N(pKp), i.e., Np(Kp) – and this provides the solution.