Libertarians, recall, are incompatibilists who think that we have free will and hence that determinism is false. The simplest kind of libertarian thinks that our acts and choices are free because they are uncaused. This is the view that Taylor calls "Simple Indeterminism."
A problem for this kind of libertarianism is the following. To say that our acts and choices are uncaused makes it sound like they are random and arbitrary. But if they are, then how could we be held responsible for them? Suppose Arnold is strapped into a machine which has a 50% chance of making him shoot a gun to the left, and a 50% chance of making him shoot the gun to the right. Surely in this case Arnold is not responsible for which way the gun shoots. He's at the whim of chance. And why should we be any more responsible if the randomness is located inside our heads, rather than in some machine that we're strapped into? Suppose there's an electric pulse travelling through my brain. When it hits a certain nerve, it will either go to the left or to the right. It has a 50% chance of going either way. If it goes to the left I will perform one act, and if it goes to the right I will perform another. I'm no more in control, in this case, than Arnold is. There's no way for me to influence the pulse, to make it go one way rather than the other. For we said that nothing causes the pulse to go in the way it does. If it goes to the left, that just happens; and if it goes to the right, that just happens, too. I have no choice in the matter. I'm at the whim of chance, just like Arnold is.
(This argument can be found at van Inwagen pp. 192-3, Taylor pp. 47-8, and Kane pp. 35-8.)
Hence, if this kind of libertarian view is correct, it's hard to see how I could be any more morally responsible for my actions than I would be if all those actions were causally determined.
In the case of an action that is free, it must not only be such that it is caused by the agent who performs it, but also such that no antecedent conditions were sufficient for his performing just that action. In the case of an action that is both free and rational, it must be such that the agent who performed it did so for some reason, but this reason cannot have been the cause of it... When I believe that I have done something, I do believe that it was I who caused it to be done, I who made something happen, and not merely something within me, such as one of my own subjective states, which is not identical with myself. If I believe that something not identical with myself was the case of my behavior---some event wholly external to myself, for instance, or even one internal to myself, such as a nerve impulse, volition, or whatnot---then I cannot regard that behavior as being an act of mine, unless I further believe that I was the cause of that external or internal event. (Taylor p. 51; see also Kane pp. 44ff.)According to agent-causation theorists, an agent's acts and choices are caused, all right, but they are not caused by any other events which took place before them. Rather, they are caused by the agent himself. He makes them occur without anything's making him do so.
This is an interesting theory. It probably better captures your intuitions about free will than any of the other theories we've looked at so far. Unfortunately, we don't really have the time or sophisticated background needed to get into it very deeply. I will just raise a few difficulties for it.
The agent-causation theorist says that whenever someone makes a free choice, he starts a new causal chain which was not present in the universe before he made his choice. Hence we can, in principle, test the hypothesis that there is agent-causation. Suppose we lock you in a room, and we have scientists monitor the behavior of all the particles and energy fields in the room, including the neurochemical activity in your brain. At a certain moment, you freely choose to raise your hand. Whoa! That should send the scientists scurrying. Because all of a sudden there are new causal processes going on in the room, which weren't caused by anything that was going on in the room beforehand. Your arm is going up, it's moving air molecules, it's reflecting light in new ways. According to the agent-causation theorist, none of these processes were caused to happen by the events that preceded your choosing to raise your arm.
Does it really seem plausible that the scientists would notice new causal processes of this sort? If so, we'll have to scrap science as we currently know it, and start over. Because science as we currently know it tries to explain how the world evolves, by treating the entire world, including our bodies, as a closed system. Scientists don't think that any "new" causal chains get introduced into the world every time an agent makes a free choice. Hence, from a scientific standpoint, the view that there's such a thing as agent-causation looks rather dubious.
When John raises his hand, science tells us that the hand motion was caused by events in John's brain. Perhaps the agent-causation theorist should accept this. He just thinks that John himself also did some causing. What is the relation between the brain events and John's alleged agent-causation? Did they both cause the hand to be raised? Wouldn't that make the hand-raising be overdetermined? Or was it rather that what John agent-caused was his brain events, and not the hand-raising itself? But intuitively, John didn't do anything to his brain. He needn't even know that he has a brain. It's not clear that there's any satisfying position here for the agent-causation theorist to occupy.
There are also more subtle philosophical difficulties with the agent-causation theory. These make it unclear whether the notion of an agent's causing something to happen, independently of all earlier events, even makes sense. Some of these difficulties are:
But now consider: did John have any choice about whether J occurred? If not, then it's hard to see how he had any choice about doing X. So John should have a choice about whether J occurred. Now does that mean that he had to agent-cause J, too? That is, he had to agent-cause his becoming an agent-cause of X? Won't the problem we're now considering then just repeat itself, at the next higher level? (van Inwagen mentions this objection on p. 194. Kane discusses it on pp. 49-51.)