Welcome back to our exploration of epistemology. In the previous lesson, we established the fundamental divide between internalist and externalist theories of justification. We saw that internalism locates justification in factors accessible to the mind, aligning with a first-person, duty-based perspective on rationality. Externalism, in contrast, insists that justification must depend on factors outside the mind, forging an objective link between a belief and the truth.
Today, we will dissect the most prominent and influential externalist theory: process reliabilism. Developed chiefly by Alvin Goldman, this theory attempts to provide a naturalistic account of justification, grounding it not in our internal states of awareness, but in the objective reliability of the cognitive processes that produce our beliefs. This approach aims to replace the often elusive internalist criteria with a more concrete, externally verifiable standard. Our goal is to understand the mechanics of this theory, see why it is a powerful form of externalism, and grapple with the significant challenges it faces.
Let's begin by situating reliabilism on the epistemological map we saw last time.
1. From Causal Chains to Reliable Processes
To understand reliabilism, it's helpful to see how Goldman arrived at it. His earlier work, the causal theory of knowledge, was a direct response to Gettier cases. The causal theory proposed that for a person S to know that p, S's belief in p must be caused by the fact that p. This elegantly handles many Gettier cases where the justification is disconnected from the truth. However, this simple causal link proved insufficient on its own.
A famous thought experiment, also devised by Goldman, demonstrated the problem and paved the way for reliabilism.
PHILOSOPHY - Epistemology: Analyzing Knowledge #3 (Causal and Reliabilist Theories) [HD]
This video from Wireless Philosophy introduces Goldman's earlier causal theory and then uses the "Fake Barn County" thought experiment to motivate the shift to reliabilism.
Watch from the beginning to the introduction of reliabilism's core idea. As you watch, focus on: The basic idea of the causal theory (the core idea). The "Fake Barn County" scenario (the story): Why does Henry seem not to know he's seeing a barn, even though his belief is true and is causally produced by a real barn? How this problem leads to the concept of reliability (the problem and core idea).
The Fake Barn County case reveals that a simple, one-off causal connection isn't enough. Henry's belief-forming process—identifying barns by sight from the road—is unreliable in his current environment. Even though he gets it right this one time, it's a matter of luck. This intuition is the cornerstone of reliabilism.
The core idea is simple and powerful: A belief is justified if and only if it is the output of a reliable belief-forming process.
A "reliable" process is one with a high truth-ratio; it produces a high percentage of true beliefs over false ones. Standard examples of reliable processes include perception in good lighting, memory, and valid deductive reasoning. Unreliable processes include wishful thinking, guessing, and hasty generalization.
2. The Structure of Process Reliabilism
Goldman's 1979 paper, "What Is Justified Belief?", laid out a more formal version of this theory. A key goal, which will resonate with your scientific background, was to provide a reductive account of justification. This means defining justification using only non-epistemic terms (like causation, belief, truth, and psychological processes), without smuggling in concepts like "evidence" or "good reasons," which are themselves epistemic and would make the definition circular.
To achieve this, process reliabilism proposes a structure that should be familiar from our discussion of foundationalism.
Reliabilist Epistemology - Stanford Encyclopedia of Philosophy
This section from the Stanford Encyclopedia of Philosophy provides a precise, formal account of Goldman's theory. It moves beyond the simple slogan to the principles that govern the theory.
Please read the section titled 2. Process Reliabilism for Justification. Focus on these key components: The idea that justification depends on how a belief is causally produced. The definition of reliability as a high truth-ratio. The distinction between processes for basic beliefs and those for non-basic (inferential) beliefs. Pay close attention to the concepts of unconditional and conditional reliability. The explanation of why process reliabilism is a "historical" theory.
As the reading highlights, reliabilism distinguishes between two kinds of processes:
- Belief-independent processes: These do not take other beliefs as inputs. Perception is a classic example. A belief formed by such a process is justified if the process is unconditionally reliable. These form the "foundations" of a belief structure.
- Belief-dependent processes: These take other beliefs as inputs (e.g., inference, memory). A belief formed by such a process is justified if the process is conditionally reliable (i.e., it reliably produces true outputs given true inputs) AND the input beliefs are themselves justified.
This historical, two-tiered structure allows reliabilism to account for the justification of complex beliefs while still grounding them in non-doxastic inputs like perception, thus stopping the epistemic regress.
3. Reliabilism as a Paradigm of Externalism
The connection to externalism should now be clear. Reliabilism holds that the justifying property of a belief is its origin in a reliable process. The reliability of that process is an objective, statistical fact about how that process type performs in the world. Crucially, the believer does not need to know, believe, or have any access to the fact that the process is reliable.
Internalism and Externalism - Epistemology Video 13
This video by Victor Gijsbers, which we used in the last lesson, explicitly frames reliabilism as a key externalist theory.
Please re-watch the segment from on reliabilism as an externalist position. The lecturer makes the central point very clearly: what matters is the objective reliability of the method, not the subject's "internal relation to that method."
This externalist nature is what allows reliabilism to straightforwardly account for knowledge in cases that are difficult for internalism. Consider the classic example of the chicken-sexer, an individual who can reliably determine the sex of chicks but has no idea how they do it and cannot articulate any reasons for their beliefs. The internalist, who requires access to reasons, struggles to say the chicken-sexer's beliefs are justified. The reliabilist has a simple explanation: the chicken-sexer possesses a reliable, albeit introspectively opaque, belief-forming process. Since the process is reliable, the resulting beliefs are justified. The same logic applies to the knowledge of young children or even animals.
4. Key Challenges to Reliabilism
Despite its elegance and explanatory power, reliabilism faces several profound objections. These challenges have driven much of the debate in epistemology for the last forty years.
a) The New Evil Demon Problem
As we saw in the last lesson, this is a major challenge. In a demon world, your envatted twin's perceptual processes are massively unreliable. According to a simple version of reliabilism, their beliefs are therefore unjustified. Yet, from the inside, their experience is identical to yours, and our strong intuition is that they are just as justified as you are. This suggests that reliability can't be all there is to justification.
b) The Clairvoyant Problem
This is the other side of the coin, challenging the sufficiency of reliability. BonJour's case of Norman the clairvoyant, who has a perfectly reliable faculty but no reason to trust it, suggests that mere reliability isn't enough. Forming a belief for no apparent reason, even if the source is reliable, seems epistemically irresponsible and thus unjustified.
c) The Generality Problem
This is perhaps the most difficult technical problem for reliabilism. The justificatory status of a belief depends on the reliability of the process that produced it. But how do we individuate the "process"? A single belief token is an instance of many different process types, each with a different reliability.
Consider Henry in Fake Barn County again. When he sees the one real barn, what process is he using?
- Process Type 1: Visual belief formation. (Highly reliable in general)
- Process Type 2: Identifying objects as barns by sight. (Highly reliable in most places)
- Process Type 3: Identifying objects as barns by sight while driving through Fake Barn County. (Highly unreliable)
- Process Type 4: Identifying this particular red object as a barn, from this angle, at this time of day. (Perfectly reliable, since it produced a true belief in this single instance)
The justificatory status of Henry's belief depends entirely on which process type we choose as the relevant one. Without a principled, non-arbitrary way to specify the relevant process type, the theory seems unable to deliver a determinate verdict.
This is analogous to a problem in experimental statistics. If you get a surprising result, its statistical significance (p-value) depends on the reference class of experiments you compare it to. Is it this single run? All runs on this specific machine? All experiments of this general type? The answer to "Is this result statistically significant?" can change based on how you define the "type" of the experiment, just as the answer to "Is this belief justified?" can change based on how you define the process type.
Reliabilists have proposed various solutions to these problems, such as adding a "no defeaters" clause to handle the clairvoyant case, or appealing to "normal worlds" to handle the evil demon. However, the generality problem remains a deep and persistent challenge.
Conclusion
Reliabilism represents a revolutionary shift in epistemology, moving the locus of justification from the internal, first-person realm of reasons to the external, third-person realm of objective reliability. It offers a powerful, naturalistic framework that aligns well with a scientific worldview and elegantly handles cases of unreflective knowledge.
- Key Takeaways:
- Reliabilism is an externalist theory stating that a belief is justified if and only if it is produced by a reliable belief-forming process (one with a high truth-ratio).
- It is a historical theory that grounds justification in the causal origin of a belief, distinguishing between unconditionally reliable processes for basic beliefs and conditionally reliable processes for non-basic beliefs.
- As an externalist theory, it does not require the agent to have any cognitive access to the reliability of their belief-forming processes.
- It faces significant challenges, including the New Evil Demon problem (reliability isn't necessary), the Clairvoyant problem (reliability isn't sufficient), and the Generality Problem (defining the relevant process type).
This concludes our analysis of the major structural theories of justification. We've examined the regress problem, the foundationalist and coherentist responses, and the deep divide between internalist and externalist approaches, culminating in reliabilism.
Preview of the Next Lesson:
Having focused on the nature of justification—what it is for a belief to be justified—we will now turn to the ethics of belief. Is there a moral or intellectual duty to believe in certain ways? We will explore W.K. Clifford's stringent principle that "it is wrong always, everywhere, and for anyone, to believe anything upon insufficient evidence," and William James's passionate pragmatist argument for the "will to believe." This will shift our focus from the structure of knowledge to the responsibilities of the knower.