Can false beliefs be useful for knowledge?

By: Michael Albalah

Note: This was originally a paper I wrote for a 400 level Epistemology class with Peter Klein. He is responsible for my introducing me to the joy of discovery and analysis necessary to appreciate western philosophy. He always gave you his time. I once said, in relation to this paper, in a tone not appropriate given my lack of exposure, "All your doing is pointing out a tiny margin of potential truth!" He looked at me and asked, "Isn't that the idea?" He spent his summers fighting forest fighters in Montana living as a Rancher. Peter Klein is a professor of philosophy and chair of the department at Rutgers University, New Jersey. He received a BA at Earlham College, and a PdD from Yale University. He is author of 'Certainty: A refutation of Skepticism." I love that. He also authored a plethora of articles addressing issues in epistemology. He is know for his work addressing skepticism. His most influential work, however, is on the nature of knowledge. He long defended the defeasibility theory yet penned his own accounts in defense of infintisism as a chain of proposition that beget justification. He has also recently advocated a picture of knowledge according to which one can have knowledge of p even if the justification n for the belief p is essentially based on false premises. Klein calls these "Useful falsehoods." I wrote a term paper arguing against his conclusion (edited and reproduced below). He gave me full credit.

The purpose of this paper is to show why false beliefs, specifically of the structure Peter Klein evisages in “Useful false beliefs” are not useful for knowledge. As structured these beliefs inappropriately allow certain propositions that utilize false beliefs to be epistemically useful for attaining knowledge. First I will do my best to represent Peter Klein’s proposed 'useful false beliefs.' I will also identify which of the examples he used that I would like to focus on. I will then articulate what I believe be the minimum fundamental nature of justification necessary for knowledge to be attained; characteristics which are not present in the Klein cases. I will then explain what epistemic luck is and how it is an inappropraite ingredient to achieve knowledge. In the cases below at leat some degree of epistemic luck contributes to the supposed knowledge-havers. Then I will explain Klein’s conjunct cases. I will also give credence to its virtues but note the threat conjunct cases pose. Finally I will propose a counter example that wouldn't intuitively be considered knowledge but still would qualify as acceptable in Klein’s structure.

Peter Klein in “Useful False Beliefs” espouses a claim that certain types of false beliefs should be counted as useful for the production of knowledge. It is important to take careful note of the requirements that are needed; there are seven of them. These characteristics expose an interesting caveat in determining what could be allowed to be a facilitator of knowledge. He claims that since a false belief can play an essential causal role in producing the cognition of the true proposition p (of certain characteristics that he describes) the one who has this type of false belief can be said to be knowledge having. His claim is that there are certain cases (he produces four examples) where some people’s intuition would be that a person uses a false belief to attain knowledge. His structure is replicated here now:

The belief that uf is a useful falsehood to S (for acquiring knowledge that h) by producing a doxastically justified belief that h iff:

1. Uf is false

2. The belief that uf is doxastically justified for S

3. The belief that uf is essential in the causal production of the belief that h

4. Uf propositionally justifies h

5. Uf entails a true proposition, t

6. T propositionally justified h

7. Whatever doxastically justifies the belief that uf for S also propositionally justifies t for S

For the sake of clarification certain terms will be described now. Propositional justification is the type of justification in a case in which a person has an “epistemically adequate basis” for the proposition h. Propositional justification differs from doxastic justification in that doxastic requires the person to actually have an “appropriate causal pedigree” to attain the belief of the proposition.

I will now quote the two cases that Klein uses in his paper that I would like to focus my contention on:

"The Santa Claus Case. Mom and Dad tell young Virginia that Santa will put some presents under the tree on Christmas Eve. Believing what her parents told her, she infers that there will be presents under the tree on Christmas morning. She knows that."

"The Ptolemaic Astronomer Case. The date is 2 September 1203; the place is Oxford University. An astronomy class is in session and the instructor, one of the most noted Ptolemaic astronomers of the thirteenth century, is showing students how to calculate the relative positions of the sun and planets both backward and forward in time using the deferent and epicycle orbits of those bodies and their (then) current positions. After carefully explaining the method, he asks the students to determine whether Mars will be visible from the earth 800 years later, supposing, of course, that it is not cloudy that night, that both the Earth and mars still exist, and so on. The Students enter the (then) current relative positions of the Sun, Mars, and Earth as they believe them to be according to Ptolemaic astronomy and then they extrapolate using the method they have just learned. They conclude that, ceteris paribus, Mars will be visible on 2 September 2003. On the assumption that the assigned orbits and then-current relative positions of the three bodies allow for sufficiently accurate extrapolations, the students know that Mars will be visible on 2 September 2003, even though their beliefs are based on false beliefs about the fixed position of the Earth and the orbits of the Sun and Mars."

I am a proponent of Reliability theories of justification. In “A Causal Theory of Knowing” Alvin Goldman provides an Invariantist account of the structure of knowledge. This means that there is a fixed set of conditions that a person must satisfy in order to know something. He argues that Smith’s proposition in the Gettier cases (which can be found in that ever prevalent paper “Is Justified True Belief Knowledge,” Gettier) the true fact that “Brown is in Barcelona” has nothing to do with Smith believing his proposition. Goldman calls for an “Appropriate Causal Connection” as a necessary condition for knowledge. He discusses certain epistemic tools that must be employed to attain knowledge. I am not going to replicate them here. What I would like to add to this account of the structure of knowledge is the requirement that the believer themselves must be of certain epistemic right to come to any conclusion about themselves having knowledge. I could be said to trust a mechanic who is reputable, that after working on the car tells me the inner chassis is behaving appropriately (even if it is not) and from that I conclude that there is at least one part of the car that is working appropriately. Having no knowledge of the inner workings of a car I cannot be said to have knowledge on the matter. Knowledge is a bold claim, one that must employ the best possible causal pattern. Not only must the actual process of attaining and transferring justification be reliable but also the state of the believer must be one that it is permissible to claim knowledge having state. The belief states have certain requirements to be allowed to be appropriate belief states that have the strength and appropriate structure to support the weight of passing justification for the purpose of knowledge. I argue that Virginia is in a realm of an “inappropriate belief state.” Since young Virginia is so gullible as to believe that Santa Claus exists and is the one that puts presents under the tree she cannot be allowed to make claims on the state of the presents at the bottom of the tree during Christmas time nor can she claim or be said to have knowledge of her proposition that someone is going to be put presents under the Christmas tree.

Epistemic luck is what is acknowledged when someone comes to a true proposition through means that were in any way accidental, coincidental, or fortuitous. According to the Incompatibility Thesis using any amount of epistemic luck is incompatible with an appropriate procedure to attain knowledge. I would like to focus specifically on an aspect that is present in every instance of epistemic luck, namely the lack of ‘control’ of the believer over the consequences. Clearly Virginia had no logical control over the process of knowing that someone was going to put presents under the tree because under her auspices it would be Santa Claus putting present and not her parents that were putting presents. She was lucky that the future events that were out of her control, as fate would have it, retroactively determined that her proposition that someone is going to put presents under the tree was accurate and knowledge inducing. Though prima facie it seems counterintuitive to claim that the Ptolemaic Astronomers could employ luck and be that lucky as to be accurate so many years later, the same argument can be said for the astronomers in the Ptolemaic Astronomy Case.

Gettier type epistemic luck is exposed when Gettier employs two principles in his cases. The two principles are the Fallibilism Principle and the Closure Principle. Fallibilism states that a person can be doxastically justified in believing a proposition p, and p can be false. The closure principle states that if a person is doxastically justified in believing a proposition p and p entails q, and that person comes to believe q on the basis of deducing it from p, then that person is doxastically justified in believing that q. The cases are too similar to the Klein cases in the following way: the supposed knowers were not in the appropriate position to make a claim on a knowledge having state (Smith does not have the Experiential Realism to make a conclusion about the whereabouts of Brown if he is to assume he has a legitimate claim as to whether he is knowledge-having about the whereabouts of Brown). Likewise Virginia’s Experiential Reality value is not at a level that one could claim knowledge. Perhaps at time ‘0’ when Virginia makes her proposition p: someone is going to put presents under her tree we would be obliged to grant her knowledge. But after careful examination of the case we must strip her of her knowledge having state. We are able to determine that she is employing a justification pattern that elucidates the fact that the conclusion was ironically an inherent coincidence. This is a unique area of epistemology that Peter Klein has discovered but one cannot say that someone so inappropriately positioned can come to claim that they have knowledge otherwise you would be guilty of attributing knowledge in cases where there was luck involved.

Earl Conee and Richard Feldman in their article entitled “Evidence” espouse the belief that evidence justifies necessarily. They call for a certain type of evidence, the specifics about the requirements to count as evidence, I do not feel is important to discuss here. What is important is to note that an evidence requirement includes what I would term “Experiential Realism.” That is to say that during the course of your life you learn that there are many things you are ignorant about. That is due to the state of your experiences or lack thereof. Clearly young Virginia has not experienced enough Christmas’s if she would so easily accept the claim that Santa Claus is going to put presents under the Christmas tree into her justification set!! Infalliblism is a theory of justification that stipulates that the justification entails the truth of the proposition for which it is justification. That is to say that the justification will never be wrong. I am proposing a justification of a weaker sort that merely depends on an appropriate causal process and a requirement that the person utilizing the justification is of a certain acceptable level of Experiential Realism. I am assuming the Ptolemaic Astronomers were the authoritative astronomers of the time and their strategies were considered wise and intuitively deemed appropriate. Perhaps it would seem then that one should grant them knowledge; had you been judging their knowledge-having-capacity at the time of their prediction in 1203. With the benefit of hindsight we are able to determine that the Ptolemaic Astronomers clearly had no clue as to the inner workings of the universe and therefore employed at least some level of epistemic luck in getting to their conclusion. They cannot be said to have knowledge. Only certain circumstances allow for a person to justify and have a claim on knowledge*.

Klein acknowledges that there could be the objector who recognizes that the false belief that is supposedly useful is really a component of a conjunct that includes the true belief that is the proposition and therefore the false belief is no longer useful but rather merely harmless. Klein acknowledges the conjunction exists but contends that this is a unique scenario and the reason why the false belief is in fact useful is because it is essential in order for the believer to have gotten to the knowledge state. It is impossible for me to not grant that. While the anonymous referee contended that a counterfactual argument is not useful in addressing the conjunct/harmless falsehood issue, I do not endorse that view. In fact I have no point of contention with Klein on this issue.

I merely wish to warn the epistemological community of the danger of what Klein’s thesis postulates. By allowing Klein’s structure you are allowing people who are clearly unqualified to claim to attain knowledge of a certain thing to claim a knowledge-having-state. Additionally Klein is allowing an unacceptable amount of epistemic luck to play a role in the knowledge having. By analogy we are looking to create a type of net that only catches the appropriate fish. By changing the net he is catching inappropriate fish (in addition to the appropriate fish) that were caught through the utilization of the new component of the net.

Counter Example:

The Egyptian God Case. Mom and Dad tell young Metit of ancient Egypt that the will of the Sun God makes the sun and the lunar cycle behave in its normal pattern (Sun rises in the East, sets in the West, and is followed by the Moon). Believing what her parents told her, she infers that the lunar cycle will behave in its normal pattern tomorrow. She does not know this.

Relevant Common Features of Counter Examples:

1. The believer (Metit) has no Experiential Realism in regards to complex nature of the premise being supposed

2. The path to knowledge includes some level of epistemic luck

*One should not assume that I am endorsing a theory of skepticism here. I believe that epistemic luck is incompatible with knowledge of the sort we are after. The Skeptic has been refuted before me by philosophical geniuses and I do not wish to attempt to recreate their arguments here, as that is not the point of this paper.