Freedom of thought, of choice, and freedom of action are necessary conditions for moral responsibility, but they are not sufficient conditions for moral behavior. Free will is a prerequisite for ethics, not the other way around.

For many philosophers, free decisions are required to be moral decisions. They fail to distinguish between ordinary responsibilities (fiduciary, financial, leadership) and moral responsibility.

This was a commonplace in ancient times. Socrates and Plato argued that "virtue is knowledge." This meant that a lack of virtue was simply a lack of knowledge. We could not be responsible for our bad actions, because we did them out of ignorance.

Aristotle disagreed. He said that our bad actions could also depend on us, even when we were doing them as a matter of habits formed long ago, as long as we are at least partially responsible for forming our habits and character.

The Scholastics thought that we are free when our decisions are rational. For them, good meant rational, so this was a variation of the ethical fallacy. We are unfree and slaves to our passions when our decisions are evil. We call this the rational fallacy.

Immanuel Kant said that we are free only when our actions are good. When our actions are bad, he said, we are slaves to our passions.

Some modern thinkers still make morality a criterion for free will, rather than freedom a requirement (some call it a "control condition") for morally responsible behavior.

Robert Kane argues that free actions, those for which we have "ultimate responsibility," must be difficult moral decisions (cf. C. A. Campbell's "effort").

Susan Wolf argues that our freedom must be "within reason" and thus free decisions are those made with full knowledge of "the True and the Good." Wolf combines the rational fallacy and the ethical fallacy.

Wolf notes an interesting asymmetry between praise and blame that echoes the ancient distinction between good and bad actions. We are quick to give and receive praise for our good actions, even when they result from luck. We tend, however, to look for mitigating circumstances for our bad actions, passing on the blame to bad luck, for example.

As early as 1890, the English philosopher Shadsworth Hodgson pointed out the confusion that freedom requires morality.

"One more remark I would make, before quitting the subject of Free-will. It is, that the kind or quality of the desires or motives, adopted or rejected in deliberation and choice, is wholly irrelevant to the question of freedom. That question concerns, not what we choose, but whether we choose at all, in any real sense of the word. Yet no doctrine is more common, especially among nominal upholders of free-will, than to represent true freedom of the will as consisting in a man's following his best impulses, obeying the dictates of his conscience, or going on to attain ever higher degrees of moral excellence or self-perfection. A great confusion of thought is here involved. Goodness of will is not the same thing as freedom of will. Its freedom is the condition of its goodness and badness alike.
(Mind, Vol. 16, No. 62 (Apr., 1891), p.179)

Hodgson is well aware of the ethical fallacy

"A power to choose only the good is a contradiction in terms; and were such a power (per impossibile) to be attained, it would be at once the highest perfection of the character, and the euthanasia of Free-will. The will would then no longer choose at all; it would have done with choosing; and the brain mechanism would thenceforward work spontaneously and habitually, no longer volitionally. The will in its new shape would indeed be free; - but free from what? From the influence of evil desires and motives, not from impediments to its power of choosing between bad motives and good ones."
(Mind, Vol. 16, No. 62 (Apr., 1891), p.179)

There is of course an undeniable historical connection between free will, determinism, indeterminism, and moral responsibility.

From the beginning of physical determinism (c. 5th century BCE), one of its proponents, Democritus, recognized that it was a threat to moral responsibility. And moral responsibility was very important to him. Nevertheless, the view of atoms and a void working by natural laws was such a gain over the traditional view of arbitrary fate and capricious gods determining our actions, that Democritus simply insisted that determinism provided humans more control for moral responsibility.

The first indeterminist was Aristotle. In his Physics and Metaphysics he said there were "accidents" caused by "chance (τυχῆ)." ² In his Physics, he clearly reckoned chance among the causes. Aristotle might have added chance as a fifth cause - an uncaused or self-caused cause - one that happens when two causal chains come together by accident (συμβεβεκός). He noted that the early physicists found no place for chance among the causes.

Aristotle's goal in the Nichomachean Ethics and Eudemian Ethics was to establish moral responsibility. He probably assumed that the human mind is somehow exempt from the materialist laws of nature, so that our actions depend on us (ἐφ ἡμῖν). In this respect, we can call Aristotle the first agent-causal free-will libertarian.

One generation after Aristotle, Epicurus (c. 4th century BCE), proposed a physical explanation for free choice as a better basis for moral responsibility. His solution was a random "swerve" of the atoms to break the causal chain of determinism, giving us more control than was possible in Democritus' strict determinism.

Epicurus wanted a purely materialist solution, one we call today event-causal libertarianism. He proposed that his random swerve could happen at any time and place. As long as there were some uncaused events in the past, there would no longer be a chain of causes back before our births. Epicurus did not want a swerve to happen at the moment of decision. That would make our actions random. But he could not explain when and where randomness could occur in his idea of free will to explain moral responsibility.

Although Epicurus' physical model for chance is ingenious and anticipated twentieth-century quantum mechanics, it provides little of deep significance for free will and moral responsibility that is not already implicit in Aristotle.

The first compatibilist, the Stoic Chryssipus (c. 3rd century BCE), strongly objected to Epicurus' suggestion of randomness, arguing that it would only undermine moral responsibility. He assumed that chance was the direct cause of action. He was also aware of the charge that physical determinism had been equated with a necessitarianism that denied any human freedom. He sought a solution to both these objections to free will and moral responsibility.