Doxastic voluntarism is the view that claims at least some beliefs can be freely chosen. It claims there are some states of affairs of belief that can be entered into by an act of the will. There are stronger views of doxastic voluntarism that claim every belief is chosen, but we need not explore that here. What follows is a brief argument in favor of the weaker version of doxastic voluntarism (DV).
1. If DV is false, then I am not rationally responsible for any of my beliefs.
2. I am rationally responsible for some of my beliefs.
3. Therefore, DV is true.
Obviously, both premises may be considered controversial by one person or another. (1) relies on the idea that free will and rationality are tied together. This is an intuitive idea. Consider a rational process; consider a piece of logical reasoning. While the process itself is considered rational, the person would not be considered rational for merely repeating the process.
Moreover, (2) assumes that humans do, in fact, possess a free will. I am rationally responsible for at least some of my beliefs. Some were formed through a process of reasoning and not a mere chemical reaction to a particular set of circumstances.
Now it occurs to me that someone may well affirm (1) and deny (2). For this objector, it is simply the case that no one ever is rational in any of his or her beliefs. But if that is true, deleterious consequences for rationality follow. Consider the following argument:
4. If no one is rationally responsible for any of his beliefs, then every belief he holds is a-rational.
5. Whatever belief is held a-rationally has no reason to be held.
6. If every belief has no reason to be held, then one has a defeater for every belief.
7. No one is rationally responsible for any of his beliefs (denial of ).
8. Therefore, every belief he holds is a-rational.
9. Therefore, every belief he holds has no reason to be held.
10. Therefore, he has a defeater for all of his beliefs.
(4) is definitional, and so it should not be denied. (5) may seem controversial, but it too is a definitional consequence (as rationality is just reasoning; if there was a reason to hold the belief as true, it would be rational). (7) is the stipulation under which the entire argument operates. (8-10) are entailed conclusions and so themselves cannot be denied. This leaves only (6).
This might seem to be a confusing premise but in reality it makes perfect sense. This is closely related to Alvin Plantinga’s Evolutionary Argument Against Naturalism. The idea is that if it is the case that some person (let’s call him Dave) has no reason to believe any of his beliefs whatsoever, then Dave cannot rationally infer that any of his beliefs are correct. Please note this is not the same as claiming that all of Dave’s beliefs are false. That is an ontological categorization, and this is concerning itself with Dave’s epistemology. If Dave cannot infer that any of his beliefs are correct, he cannot discern which beliefs are correct and which are false. If he has no reason to believe any of his beliefs are true, then, if Dave were able to be rational, he should have a defeater for every belief (since every belief would be just as probably false as true and lack any reason for believing them). But then we see that the following argument holds:
- If DV is false, then I have a defeater for all of my beliefs.
- If I have a defeater for all of my beliefs, I cannot believe (11).
- I can believe (11).
- Therefore, I do not have a defeater for all of my beliefs.
- Therefore, DV is true.
The next post will deal with practical applications of DV in apologetics and evangelism. This is a necessary primer.
 As an example, consider a computer. People may, anthropomorphically, refer to a computer as “thinking” or “reasoning,” but people recognize computers do not, in fact, reason as humans do (this is why experts in the field are desperately searching for AI).
 Of course, someone may rightly point out that (2) does not require, on its own, the idea that free will is central to rationality. That is, someone can affirm (2) but deny (1).