Believing, just believing

The question of the day is: What is a justified belief? In other words, what are we entitled to believe? The answer of the day, and the subject of my analysis, is the properly basic belief, a central concept of foundationalist theories of knowledge and rationality. The conclusion of the day is that the so-called properly basic beliefs are deeply problematic, and that personally, I would not even use the term. Properly basic beliefs as an epistemic category are of special interest to the philosopher of religion, because they have been invoked as a means to defend the rationality of religious beliefs, or at least the belief in a God. This has been part of the strategy of reformed epistemologists, of whom Alvin Plantinga is probably most well-known. Before explaining the use of the concept in reformed epistemology, which represents a modest form of foundationalism, it is appropriate to lay out the principles of foundationalism in general.

Foundationalism originated as an early modern development in epistemology. After the long period of religiously grounded, modest confidence in human reason in the Middle Ages, philosophers, beginning with René Descartes (1596–1650), engaged in the search for absolutely certain knowledge. They found it, as it were, in various ideas, beliefs or categories that were seen as foundational or logically prior to all other knowledge in an individual human mind. Descartes had his ”clear and distinct ideas”; Immanuel Kant (1724–1804) discovered a priori categories of understanding. Generally, foundationalism treats human knowledge as consisting of two kinds of beliefs: basic ones and derived ones. The former, accepted without being supported by other beliefs, constitute the ground on which the structure of knowledge is built. Beliefs can be called ”properly” basic if it is reasonable to accept them independently of other beliefs. An example of improperness would be, for instance, believing that someone is guilty of a crime just because one dislikes that person.

Strong foundationalism is a notorious position where only self-evident or incorrigible beliefs can be accepted as properly basic. Such could be arithmetical truths and beliefs concerning one’s own, immediate experience, for example. It is deemed rational to hold them and beliefs logically derived from them. Strong foundationalism, however, has been met with devastating criticism, for it is actually self-referentially incoherent. The central requirement of strong foundationalism can hardly be a self-evident or incorrigible belief, and so it cannot provide a secure basis for rational beliefs. In other words, it is not rational for a strong foundationalist to be a strong foundationalist.

Reformed epistemology has been developed with the pitfalls of strong foundationalism in mind. Its advocates, such as Plantinga, accept a modest form of foundationalism. They do think that beliefs are either basic or derived, and that any system of rationally justified beliefs requires properly basic beliefs as its foundation. The difference to strong foundationalism is that we can consider any of our beliefs as properly basic if we don’t think we have derived them from other beliefs. The reformed epistemologist does not want to overproblematize knowledge: she is confident that human structures of belief and knowledge are basically in order, and that we do know most of those things that we think we know.

For Alvin Plantinga, this broad criterion means that the belief in God’s existence can be properly basic. He has been criticized for taking this stance – for example, Gary Gutting has objected that only such beliefs should be considered properly basic of which there is no widespread disagreement. Plantinga, however, rejects consensus as a criterion for proper basicality: ”[T]here is no reason to assume, in advance, that everyone will agree on the examples [of properly basic beliefs].” (Reason and religious belief, Peterson et al. 2003, p. 114) Why should a Christian conform to the examples of, say, an atheist? Basicality is affirmed or denied according to a belief’s provenance, not to its reception, and so everyone can legitimately have their own basic beliefs.

This person-relatedness inevitably paves the way for a kind of relativism. Philip Quinn has expressed concerns over this: in his view, it is problematic that anyone can claim even the most eccentric beliefs to be properly basic. Is one not, in defining proper basicality, dealing with matters of truth or even plausibility? This is a good question indeed. But in my view the big problem is that, in construing the notion of proper basicality, one limits the range of epistemology in a way that leaves us with a truncated, implausible, and not very interesting conception of what human knowledge is. I now move on to explicate what this means.

Reformed epistemologists and other foundationalists operate in a world of beliefs that are expressed as grammatically correct sentences. Consider the following example, given in Reason and religious belief: ”[Y]ou walk into a friend’s house and, because of what you are hearing, form the belief that someone is playing the saxophone in the next room. This is a basic belief, not inferred from other beliefs of yours … you just hear the sounds and find yourself with the belief that there is a saxophone in the next room.” (p. 113) Here, the foundationalist refuses to think further where it is both possible and necessary to do so. The belief in question does, in fact, logically entail a number of prior convictions. You have to believe that there is a musical instrument known as the saxophone, and that a player is required for the instrument to produce sounds, and even that the next room actually is there!

I stress that none of these prior convictions has to be part of a conscious process of reasoning that eventually produces the belief ”someone is playing the saxophone in the next room”. But logically, they have to be there. You do not form the belief exclusively ”because of what you are hearing”. If you did, probably any living creature capable of hearing and seeing could form the same belief as you do, given the same sensory input. But obviously, animals could not. Neither could all of your fellow human beings, especially not those who had never got to learn anything about the saxophone. The sensory input in question could not possibly cause such people to believe that someone is playing the saxophone. At best, they could believe that someone is playing a musical instrument. This I take to be self-evident, but the foundationalist seems unable to account for the fact that, even upon observing the same event, different people may form different beliefs.

A more detailed description of belief formation could be the following. What you are hearing interacts with your prior convictions, whether they be conscious or subconscious, and from this interaction emerges the belief – it is an interpreted experience, achieved through reference to some form of prior knowledge. In this case, reference is made to some things you have learned about the saxophone and the structure of your friend’s house. I label these things resources of interpretation. Or I could say they are implicit truth claims inherent in the explicit truth claim that is your belief. It is downright misleading to describe the situation as if the external stimuli had immediately caused the belief to appear in your mind as a grammatically correct sentence. It just does not happen.

Furthermore, it is far from self-evident that the sphere of correct language encompasses all human knowledge. No perfectly formed sentences pop out of nothing when a child learns to communicate. Prior to sentences, single words are uttered – how come they could not stand for knowledge, beliefs, or proto-beliefs if you like? It is entirely reasonable to maintain that, for instance, if a child utters the word ”mama” when approached by its mother, it is expressing knowledge of or belief in something. Neurologist Antonio Damasio has even argued that, in a conscious mind, there can be concepts even when there is no capacity for linguistic expression whatsoever, and that concepts logically precede words. It is, I understand, of reasons of convenience that philosophers like to discuss well-formed statements, but I wouldn’t deny that knowledge is manifest even elsewhere in human communication.

I maintain that beliefs necessarily come about as interpreted experiences, through reference to some resources of interpretation. Thus, if one wants to find out whether a belief is justified, one will have to inquire about the epistemic justification of the resources involved in the formation of that belief. For example, a person who believes in God could ask: ”Was it justified to possess the resources of interpretation that allowed me to form the belief that there is a God?” When the resulting belief is a basic one, the resources of interpretation have to be subconsciously held convictions, states of taking-something-to-exist, taking-some-conditions-to-prevail. They cannot be premises of a conscious deductive process, for if they were, the belief would not be a basic one. But is the question of justification even relevant when discussing those subconscious resources? Isn’t it so that one just happens to have certain resources, and that there is no inquiring about their justification at all?

If resources of interpretation were seen as implicit truth claims underlying an explicit one (i.e., the eventual belief), one could try to reconstruct belief formation as a deductive process. The resources of interpretation being turned into premises of deduction, their status as justified and reasonable beliefs could be discussed. In effect, though, this move would undermine the concept of basic belief, for the premises, as explicit truth claims, would be no more self-evident than the conclusion. One would have to find out what implicit truth claims lie hidden in them, and repeat the procedure. Logically, an infinite regress would follow – as a reconstruction of how the eventual belief came about! This all goes to show that the notion of basic beliefs implies an a priori confident attitude towards the subconscious resources involved in belief formation.

Alvin Plantinga has exemplified this central role of confidence in a specific case: supposing God exists, it would be reasonable to think that the cognitive mechanisms producing the belief in God work reliably. Plantinga posits the sensus divinitatis (sense of the divine) as a component in human cognitive makeup intended to produce in us the belief in God. In doing this, he wants to show that under certain conditions, which actually might prevail – since the existence of God has not been disproved – the belief in God could be warranted, i.e., point to a fact. But he admits that if there is no God, then we have no sensus divinitatis and the belief in God is probably not warranted. So it seems that, at least in the God-case, the resources of interpretation are considered reliable on the assumption that the eventual belief points to a fact. The belief is justified insofar as it points to a fact. But since you can’t know whether it does, you can’t know whether the belief is justified. You are believing, just believing – you are something like a fideist, are you not?

The problematics discussed above makes it evident that foundationalism in its different forms has to be a clearly demarcated epistemology, focusing exclusively on beliefs expressed as truth claims. Any other forms of knowledge, such as subconscious resources of interpretation, are outside its scope. But it is clear that other forms of knowledge are also epistemically relevant. If foundationalist theories can handle only a specific kind of knowledge, how relevant are they really when considering human knowledge in its full diversity? Do they explain anything – do they justify anything? I have figured out two attempts to justify the information which, of necessity, precedes the formation of any belief, even a ”basic” one. The first way led to postulating an infinite series of underlying beliefs for any actual belief, and the second one to justifying a belief on the sheer assumption that it is true. Neither one of the solutions is satisfying. But can the problem of justification hereby be declared a non-issue? I will continue on this in a further post, perhaps even the next one.

Further readings:

Peterson, Michael et al.: Reason and religious belief. 2003, Oxford University Press.
(Specifically chapter 6, Knowing God without arguments: Does theism need a basis?)