With freedom comes responsibility. We are responsible for the choices we make but not for events which are outside our control.
The “Principle of Alternative Possibilities” asserts that “a person is morally responsible for what he has done only if he could have done otherwise”.
Another definition is that “a person is morally responsible for what he consents to”.
Generally we can assume others have free will, whether or not they actually do have it, without any contradictions emerging. However, practical problems start to emerge as soon as we enter the area of justice, praise and blame, reward and punishment.
It is not fair or “just” to reward or punish actions which are not under our control. Applying such measures may have benefits in changing the behaviour of the recipient and of others looking on but to use the word “justice” is no more appropriate then applying it to correcting or improving the performance of a machine. There is no intent involved, just automatic responses.
This can be seen when we consider people who, for whatever reason, may be considered to be less culpable because of “diminished responsibility”. They are treated more as automata and less as free agents and the punishment, or therapy, which is meted out reflects this.
In practice, the main determinant is the pragmatic result of reward and punishment in controlling behaviour regardless of its justice or its fairness. A person's responsibility for their actions is then likely to be defined by whether, or not, their behaviour can be changed by reward or punishment.
Hard determinism implies that moral responsibility does not exist for a person any more than it does for a machine. We are merely observers as the laws of physics play out. It would make no sense to say we were responsible for what happens any more than to say that the sun is responsible for shining.
Libertarianism and compatibilism would draw a distinction between actions for which an agent is responsible and those for which he is not. We react to a person’s action differently depending on whether we believe it was freely chosen and, if it was not, then according to the nature of the constraint or the duress which was applied.
Even if we are not just our brains and bodies, we do rely on these for interacting with the world in a similar way that a car driver relies on the car for transport. If there is an accident then it could be down to driver error, for which he would be responsible, or mechanical failure of the car, for which he might not.
Likewise, a particular behaviour might be the responsibility of the person if it was chosen and consented to, alternatively it could be due to a brain malfunction or due to receiving false information.
How one tells the difference is problematic. It depends on what we believe constitutes the “driver” and what constitutes the “car”. This has been the subject of many arguments about mitigation by “diminished responsibility” in courts of law.
For both the Libertarian and the Compatibilist viewpoints, the unknown is where to place the boundary between the person, as an agent, and the external world. For instance, is it our eyes? our optic nerve? The V1 part of the brain, some other part of the brain? Or a separate entity which is connected to the brain in a similar way that a driver is connected to a car?
Since this is unknown and is the subject of shakily founded beliefs, the way we treat other people is more to do with either producing a desired change in behaviour or to an emotional response to being wronged or blessed by others, regardless of justice or truth.
The tangles we get caught up in when we consider moral responsibility are well illustrated by the song in West Side Story performed in the video below.