Primary Source: Epistemology: A Contemporary Introduction to the Theory of Knowledge by Robert Audi
Humans have two different kinds of beliefs. The first kind of beliefs is perceptual. An example of a perceptual belief is, “I hear chirping outside the window.” The second kind of beliefs is inferential. An example of an inferential belief is, “Birds are outside the window because I hear chirping.” Inferential beliefs are called as such because they take perceptual beliefs and use logic to infer another belief. Epistemology is interested in knowing what justifies these beliefs. Foundationalists argue that even if “there could be infinite or circular beliefs chains, they could not be sources of knowledge or justification.”(Audi 205) The foundationalists claim that knowledge and justification for beliefs are both possible by foundational beliefs. These foundational beliefs are non-inferential and perceptual in nature.
Beliefs Through Experience
There are four basic sources of knowledge or justified beliefs. The first source of knowledge is perception. A perceptual belief is a belief known through the senses. For example, “I perceive through my senses that a chirping sound is coming from outside the window.” The second source of knowledge is consciousness. The example, “I am thinking about X” best illustrates this source of knowledge/justified belief. The third basic source of knowledge or justified belief is reflection. Robert Audi, a foundationalist, describes the following as an example of reflection: “if person A is older than B and B is older than C, then A is older than C.”(Audi 205) The last type of basic knowledge/justified beliefs is memory. An example of a belief based on memory is, “I recall that five minutes ago I heard chirping outside the window.” So, to the foundationalist, these types of experiences bring about knowledge or justified beliefs.
An epistemic chain “is simply a chain of beliefs, with at least the first constituting knowledge, and each belief linked to the previous one by being based on it.”(Audi 206)
4 types of epistemic chains:
Foundationalists reject epistemic chains 1 through 3 and argue that the 4th is valid.
- Justified beliefs/knowledge can come from epistemic chains that are not based on direct perceptual beliefs.
- Justified beliefs/knowledge must come from epistemic chains that are based on direct perceptual beliefs.
Infinite Epistemic Chains: Concepts that seem infinite are really just ungraspable to human. An example Audi uses to illustrate the ungraspable infinite is based in arithmetic. Take the number 1 and multiply it by 2, equaling 2. Then take the number 2 and multiply it by 2, equaling 4. Take the number 4 and multiply it by 2 equaling 8 and so on. At some point, the numbers will begin to get ungraspable. These ungraspable numbers are not infinite.
Circular Epistemic Chains: These chains beg the question. Inferential beliefs are drawn that lead right back to the original premise.
A -> B -> C -> A
Justified Epistemic Chains That Are Not Based On Direct Perceptual Beliefs:
An example of this epistemic chain can be illustrated in the following example:
I hear a bird chirping outside.
I believe I know that my neighbor’s car will breakdown today.
My neighbor’s car did break down today.
Is the belief, “My neighbor’s car did break down today,” justified knowledge? This statement would be based on “inclinations” that it would come true. These “inclinations” are not based on the bird chirping. The belief about the car breaking down would not be considered justified knowledge, but it would be considered close to it. There must have been a reason for believing that the car would break down. This helps the foundationalists claim because at some point, the knowledge of the car breaking down would be justified through perception and not bird chirping (even though bird chirping was given as justification).
Justified Epistemic Chains That Are Based On Direct Perceptual Beliefs: The only epistemic chain that allows for foundationalism is this option. The basic beliefs we have through experience (perception, consciousness, reflection, & memory) will grant us direct knowledge that can be used to infer other beliefs. Those other inferred beliefs like, “A bird is chirping outside the window,” is justified by the direct knowledge that, “I hear chirping outside the window.”
The following illustrates inferred beliefs based on direct knowledge:
X: I hear chirping outside the window.
Y: A bird is outside the window.
X is a foundational belief. X is a foundational belief because it is known though the senses (direct knowledge). Y is inferentially based on the foundational belief in X.
The argument for the 4th epistemic chain, also known as the foundationalists regress argument, appears as follows:
If knowledge is attainable then it must occur in an epistemic chain.
The 4th option is the only possible epistemic chain.
The 4 types of epistemic chains are mutually exclusive meaning that it is not possible for more than 1 to occur.
For knowledge to exist it must occur in the 4th option which claims that we can have direct knowledge.
Therefore, if knowledge is possible to attain, then some of that knowledge must be direct from experience (perception, consciousness, reflection, memory)
Analysis Of The Foundationalist Regress Argument
The conclusion of this argument supporting the 4th type of epistemic chain anticipates a skeptic’s response. If the conclusion of the argument was “Therefore we do have direct knowledge,” then a skeptic could apply a global skeptical hypothesis like Descartes’ evil demon. A possible skeptic response to this modified conclusion would be, “how do you know that direct perceptual knowledge is not being distorted by an evil demon?” The conclusion of the foundationalist regress argument, step 5, was left as a conditional, which allows for the possibility of no knowledge existing.
From the foundationalist regress argument one could conclude that knowledge can be traced back to non-inferential claims based in experience. The argument states that if there is inferential knowledge then there is non-inferential direct knowledge. By this, we can deduce that knowledge only occurs in foundationally grounded epistemic chains even if the same belief is in another circular/infinite chain. For example, lets look at a grounded epistemic chain and use the grounded inference in non-grounded chain.
Grounded Chain: A -> B -> C
C is justified inferentially.
Non-Grounded Chain: X -> Y -> C -> X
C is not justified inferentially.
The last point that needs to be made about the conclusion of the foundationalists regress argument is that it does not matter what beliefs a person holds as long as they are grounded. The argument deals with the structure of attaining knowledge rather than claiming that individuals should hold certain beliefs.
Support for Fallibilist Foundationalism
Fallibilist foundationalism is a moderate version of foundationalism. Foundationalism can be fallible because knowledge does not inferentially force that “someone’s grounds for knowledge are indefeasible. One can fail to know a proposition not because it is false, but because one ceases to be justified in believing it.”(Audi 209) The fallibilist argument is as follows:
The structure of justified beliefs is foundational and therefore non-foundationally justified beliefs are inferentially justified.
The justifications of beliefs have the possibility of being defeasible.
The inferential transmission of justification does not need to be deductive rather it merely needs to be probabilistic.
Therefore, the justification of non-foundational beliefs does not need to deduce all justification from foundational beliefs rather non-foundational beliefs must be justified enough so that they would still be “justified if any other justification they have were eliminated.”(Audi 208)
There are five reasons to support the conclusion of the fallibilist foundationalist. First, the argument appeals to a possible skeptic response, using the same logic as the foundationalist regress argument. Secondly, the fallibilist argument appeals to common sense by stating that individuals are justified in holding foundational beliefs that people should hold. It makes sense for people to believe what they perceive through their experience and therefore, individuals are justified in those beliefs. Third, psychology supports fallibilist foundationalism through naturalistic experimentation of human development. These studies done by psychologist show that human beliefs are learned both from experience and inference. Fourth, the reason why people hold perceptual beliefs is because sensory experience justifies them. When looking at evolution, non-inferential beliefs are essential to our survival as a species. Fifth, the fallibilist foundationalism is not dogmatic because it allows people to have justified beliefs based on their perception and experience. If the human experience changes over time, then our beliefs would also change.
Laurence Bonjour’s Epistemic Ascent Argument Against Foundationalism
Laurence Bonjour is a coherentist who claims that justification for foundational beliefs must come in the following form:
A certain foundational belief has property P.
Beliefs that have P are likely to be true.
Therefore, foundational beliefs that have P are likely to be true.
Bonjour argues that the only way to have a basic foundational belief is to stipulate that it exists. In other words, he states that foundational beliefs are foundational because “I said so.” He claims that there are no self-justified beliefs rather beliefs, including foundational ones, ought to be justified by an argument. If all beliefs were justified by an argument structure, like the epistemic ascent, then there would be no basic foundational beliefs because they rely upon non-inferential perception. Therefore, Bonjour leaves us with coherentism, which prefers epistemic chains in the form of loops.
Fallibilist Foundationalist Response To Bonjour’s Epistemic Ascent Argument
D.M. Armstrong, a foundationalist, developed a response to Bonjour’s epistemic ascent argument. Armstrong argues that there is no need to have cognitive possession for as to why beliefs with property P are likely to be true. “What makes a true non-inferential belief a case of knowledge is some natural relation which holds between belief-state… and the situation which makes the belief true.”(Armstrong 217) Justified beliefs depend on a relationship between the external outside world and the perceptual view of that interaction. No more justification is needed for direct perceptual beliefs. Furthermore, if cognitive possession were held then a more “rudimentary” state would be revealed instead of the coherentist claim that more beliefs would be exposed. Therefore, the coherentist view is mistaken in claiming that there are no directly perceived foundational beliefs.
Fallibilist foundationalism clarifies our understanding of knowledge and how beliefs are justified. Through epistemic chains based on direct perceptual beliefs, we are able to foundationally ground beliefs and inferentially justify non-foundational beliefs. Fallibilist foundationalism adds that not all non-foundational beliefs need to be deductively justified in foundational beliefs rather non-foundational beliefs must be justified enough so that they would still be “justified if any other justification they have were eliminated.”(Audi 208)