This was written in response to a comment the Dane made over here. The topic was Plantinga's criticism of classical foundationalism. The question was, "Well, taking Plantinga's criticism of foundationalism to his own thoughts, how can belief in God (or in his actions) be properly basic - since so many vary on whether God exists or not?"
Plantinga's critique of classical foundationalism (CF) for "its excessive exclusion and its self-referential incoherence" cannot be applied to his idea that belief in God is properly basic. According to CF, a belief, in order to constitute knowledge, must either be properly basic or based on propositional reasoning from a properly basic foundation. There is nothing wrong with the idea that certain known things follow this paradigm: foundational propositions + reason = knowledge. The weakness of CF is found in its self imposed limitations. A properly basic belief in CF must either be self-evident for me or immediately about my experience (Warrant and Proper Function p.182). A properly basic belief of the second kind says nothing about the external world. Therefore, the proposition "I see a black dog" is not properly basic. The properely basic belief in this case is, "I am being appeared to both blackly and dogly." It is only through a relatively complex reasoning process that I can be justified in saying that my experience results from the existence of a black dog external to my own mind. The self-referential incoherence of CF comes in the fact that the belief "a properly basic belief must either be self-evident for me or immediately about my experience" is itself neither self-evident for me nor immediately about my experience; nor does it have such a belief as a premise.
By accepting the idea of properly basic beliefs, Plantinga demonstrates that he is a foundationalist. He rejects the other two options, which are 1) Dismissing altogether the possibility of epistemology and, consequently, ability to know the truth. This is repugnant to Christian claims, especially Reformed claims, that knowledge is possible. 2) Coherentism. According to a popular misunderstanding of this system (unfortunately, even by some who hold to it), coherentism is a big epistemic circle. Belief A is based on belief A(1), which is based on A(2), and so on to A(n), which is based on A. Warrant for belief is being transferred all over the place, but there is no accounting for how the warrant got there. Transfer does not equal creation. A more likely coherentism does not claim that warrant is created in the process of transfer, but that the fact of systemic coherence itself constitutes warrant. Plantinga disproves this view by his telling of the Epistemically Inflexible Climber. Briefly, a climber scales a mountian and sits on a ledge. He forms certain beliefs about his situation. He is sitting on a mountain, wearing red shoes, the sun is shining, an eagle is circling below, etc. All of these beliefs are coherent- they do not contradict one another. Suddenly, a rare burst of solar energy damages a part of his brain causing his latest beliefs to become fixed. Even when his friend manages to get him down the mountain and takes him to the opera in an attempt to cure him, he still believes that he is on a mountain, in red shoes, watching an eagle, etc. His beliefs are just as coherent as they were before, but they hardly match the reality of his present situation.
Coherentism is fine as far as it goes- we want our beliefs to match up with each other. But there needs to be something more. This can only be found in some form of foundationalism. Certain beliefs need to be properly basic, not in the limited sense of CF, but in the sense that they have an immediate correspondence to external reality. A properly basic belief does not need to rely on any self-evident propositions, nor any propositions at all, for that matter. I would be warranted in my belief that I see a black dog simply by virtue of the fact that I see a black dog. There is no need to go through a reasoning process in order to be justified in concluding that there really is a black dog. I am warranted in believing that there is a black dog on the basis of seeing a black dog provided that two things are true; 1) there really is a black dog and 2) my senses and mental faculties are functioning within normal parameters. However, and this is a key difference between CF and this broader variety, I don't have to form any beliefs at all about these conditions for knowledge in order for warrant to obtain. Whereas CF is a form of internalist epistemology, in which I am constantly asking, "how do I know that I know?" this other foundationalism is an externalist epistemology. It seeks, not to eliminate all doubt as in the Cartesian model, but only to explain the mechanism of knowledge. It assumes the existence of a creator. [This is not to say that belief in God is necessary in order for knoweldge to be warranted, only his actual existence.] We are placed in a world of external objects and are given both mental faculties and sensory abilities designed to give us knowledge of the existence of these objects. The difference lies in a paradigm of faith as opposed to one of doubt.