Are Your Belief-Forming Faculties Reliable?

Elliott Crozat


May 27, 2020

In Plato’s Meno, Socrates and Meno discuss the nature of knowledge. Meno asks the master of dialectic about the difference between knowledge and true belief. Meno wonders if there is a real difference, and assuming there is, why knowledge is better. Socrates responds by comparing true belief to one of Daedalus’ statues.[i] The statues are beautiful, Socrates says, but more valuable if tethered. Similarly, a true belief is good; but it falls short of knowledge. Knowledge has greater value because it is rationally grounded, whereas mere true belief is ungrounded.

According to Socrates, knowledge is true belief plus a rational account (logos) to ground the belief.

“True opinions are a fine thing and do all sorts of good so long as they stay in their place, but they will not stay long. They run away from a man’s mind; so, they are not worth much until you tether them by working out the reason… Once they are tied down, they become knowledge, and are stable. That is why knowledge is something more valuable than right opinion. What distinguishes one from the other is the tether.”[ii]

In short, according to Plato/Socrates, knowledge is justified, true belief (JTB). I.e., for any subject S, S knows that p iff:

  1. p is true;
  2. S believes that p;
  3. p is justified for S.

Thinking About Justification

Plato’s view of knowledge as JTB was generally accepted until Edmund Gettier challenged it in his 1963 paper entitled Is Justified True Belief Knowledge?[iii] Since the publication of Gettier’s paper, most epistemologists agree that Plato’s view needs further support. It seems Plato was right that knowledge requires justification, truth, and belief. However, the nature of justification is unclear, and there might be some fourth condition necessary for knowledge. Plato himself was wary that knowledge is only JTB. Epistemologists have thus focused on the nature of justification to bolster the JTB analysis of knowledge.

Internalism and Externalism

The post-Gettier emphasis on justification has generated a debate between internalists and externalists. Roughly, internalists hold that justification is wholly a matter of evidential factors internal to a person, such as beliefs, memories, and experiences. Externalists disagree, asserting that justification requires factors external to a person.


As a theory of justification, reliabilism states that justification is determined by a reliable psychological process; i.e., one’s belief is justified iff the belief was produced by a trustworthy process. The factors of such a process are generally external to the subject. For this reason, reliabilism is a version of externalism.

Reliabilism has several virtues. First, it provides an apparently plausible answer to Gettier’s challenge, especially concerning the problem of epistemic luck. Second, arguably, it enables one to hold that S’s belief is justified without requiring that S be infallible. Third, it is a reasonable response to the sort of skepticism which holds that knowledge requires objective certainty (i.e., infallibility). Reliabilism is a fallibilist approach to epistemology: fallibilism in one form or another has strong appeal and is widely accepted, allowing epistemologists to hold that human beings – though susceptible to intellectual error – possess various items of knowledge.[iv] Fourth, its externalist nature is thought amenable to investigation by cognitive psychology and neuroscience.

However, there are multiple concerns about reliabilism. I will briefly address four. First, reliabilism appeals to a passive process of belief-production which seems to exclude factors such as freedom of the will, intellectual responsibility, deliberative control, and rational autonomy.

For example, some worry that if one does not possess freedom of the will in a libertarian sense, then one’s beliefs are not always formed via a reliable process. Tim Stratton provides the following argument to make this point.

The FreeThinking Argument

  1. If humans are not free (in the libertarian sense), they cannot either rationally infer or rationally affirm claims of knowledge.
  2. Humans can rationally infer and rationally affirm claims of knowledge.
  3. Therefore, humans are free (in the libertarian sense).

To defend this argument, Stratton notes that the second premise seems uncontroversial and one that most will affirm. After all, if one rejects this premise, one rejects his own ability to infer or affirm that this premise is false. So, the first premise is the key premise. But is it controversial? Stratton has defended the key premise by simply pointing out that it communicates the following:

“If something outside of human control causally determines you to affirm a false belief, then it would be impossible for you to infer or affirm a better belief — let alone the truth!”

Stratton points out that if naturalistic determinism is true, then all human thoughts and beliefs are causally determined by the forces of nature, the initial conditions of the big bang, past events, perhaps some quantum mechanics, and all other things outside of human control. Since these naturalistic events are not intentionally aimed at truth, then human thoughts and beliefs are not reliable in the sense of being truth-conducive (this is especially the case when it comes to metaphysical views). 

Stratton continues to explain that theological determinism has similar problems regarding reliability (See, A Rational Refutation of Divine Determinism). After all, if exhaustive divine determinism is true, then God always causally determines all thoughts and beliefs. Initially, one might think this to be unproblematic since God is an intentional agent who has the ability to “aim at truth.” The problem, however, is that human beliefs are not aimed at truth, but rather, God’s will. Given the preponderance of a plethora of disagreeing views about theological issues (even amongst Christians), it stands to reason that God’s will is not necessarily for humans to affirm true theological beliefs — this also includes the beliefs regarding the issue currently under discussion. Thus, theological thoughts and beliefs (not to mention a multitude of other thoughts and beliefs regarding other issues) are not reliable in the sense of conducing to truth. On exhaustive divine determinism, God causally determines many people — all people — to affirm false beliefs. 

The FreeThinking Argument provides a forceful defeater against a “passive external reliability” and the importance of being able to think and deliberate in a libertarian sense. However, if one is not convinced by this argument, one might simply claim a strong intuition that justification requires one or more of the factors mentioned above (namely, freedom of the will, intellectual responsibility, deliberative control, and rational autonomy). The arguments for reliabilism are insufficient to outweigh this intuition.

Additional Challenges

Second, consider the so-called “value problem.” Plato recognized that the difference between a true belief and a JTB is at least partly a matter of value. A JTB is more valuable than a mere true belief, and this value is due to the justification condition. However, reliabilism fails to explain why justification is valuable. On reliabilism, justification occurs by an external, belief-producing process which happens to a subject, perhaps in a manner that is completely beyond his ken. It is unclear how such a process answers the value problem. For example, some epistemologists hold that a JTB is valuable only if the justification is developed by the subject in a responsible or virtuous manner such that the subject deserves credit for due diligence.[v]

Third, consider what I will call here the genie problem. Suppose Smith suffers from routinely committing the wishful thinking fallacy.[vi] One day, to his great delight, Smith finds a magic lamp. He uses it to access the genie inside. The genie grants Smith the following wish: whatever Smith wants to be true is true in virtue of his wanting it. In other words, whenever Smith engages in wishful thinking, that session of wishful thinking produces a belief and makes it true. Therefore, wishful thinking is a reliable process for producing true beliefs. But this is absurd. Wishful thinking is a fallacy, not a justification. Fallacies are mistakes in reasoning, not successful efforts at justification.

Fourth, there appears to be a sorites problem.[vii] According to reliabilism, one’s belief is justified iff produced by a reliable psychological process. But what exactly is a reliable psychological process? The standard answer is that a reliable process is one which is truth-conducive. But what does it mean for a process to be truth-conducive? Again, the standard answer is that the process is truth-conducive insofar as it usually produces true beliefs. Now, consider a further question: how often must a process produce true beliefs in order to qualify as truth-conducive? 100% of the time? That would reduce reliabilism to a form of externalist infallibilism, which undercuts the reliabilist’s motivations to avoid infallibilism and skepticism. 99% of the time? That seems too strict. 75% of the time? That seems arbitrary. Why exactly 75%? Why not 74%, or 76%, or some other frequency? Indeed, any specific frequency greater than 50% and less than 100% seems arbitrary.


Reliabilism is an interesting theory of justification that has been defended by influential philosophers such as David Armstrong and Alvin Goldman. There are some motivations to accept the theory. But there are several objections to it. The debate remains open. This sketch merely outlines its main points.


[i] In Greek mythology, Daedalus is a sculptor and architect. It is said that his sculptures were so life-like that they had to be tethered to the ground to prevent them from running away.

[ii] See Plato’s Meno in Plato: Collected Dialogues, 20th Edition. Eds. Edith Hamilton and Huntington Cairns. (Princeton University Press, 2009), 381-382. See also Plato’s Theaetetus, in which Socrates characterizes knowledge as true belief supported by reasonable account.

[iii] Gettier’s paper is available at

[iv] Roughly, fallibilism is the epistemological view that human beings are vulnerable to intellectual error and thus not intellectually infallible. Consequently, for every human belief, it is possible that the belief is false based on some error made by the believer (unrestricted fallibilism); or, for some but not all human beliefs, it is possible that those beliefs are false based on error (restricted fallibilism). Hence, human beings lack objective certainty in all (or some) cases. This is a just a sketch of fallibilism. Challenges arise when trying to adjust for various factors, such as necessary truths. See

[v] It should be noted that some virtue epistemologists are also reliabilists. This view is called virtue reliabilism. Virtue reliabilism has a plausible answer to the value problem, though virtue responsibilists might disagree. Virtue reliabilism does not rule out libertarian freedom, because – on virtue reliabilism – the reliable process which justifies one’s beliefs can be developed by freely choosing to develop and maintain one’s cognitive virtues and abilities. See

[vi] See

[vii] See


About the Author

Elliott Crozat

Elliott R. Crozat is a full-time professor of philosophy and the humanities at Purdue University Global. His philosophical interests include metaphysics, epistemology, philosophy of religion, ethics, and the meaning of life. He lives in Sarasota, FL.

Learn More

More from this author