Critically assess the claim that one knows that it is raining if and only if one has a true and justified belief that it is raining.
In this essay I will critically examine the strong claim that one knows that it is raining if and only if one has a true and justified belief that it is raining. By examining both the necessity and sufficiency of the Justified True Belief (JTB) conditions, I will argue that while justification, truth, and belief are individually necessary for knowledge, they are not jointly sufficient. This insufficiency, highlighted by Gettier-style counterexamples, stems from the fallibilist nature of justification and reveals a fundamental limitation in the traditional JTB account of knowledge.
To assess this claim, let's first examine the necessity of each condition in the JTB account.
Knowledge requires a cognitive stance or commitment toward the proposition. Without belief, the individual is not mentally committed to the truth of the proposition, thus knowledge cannot be attributed to them. For example, if a person says, "I know it's raining, but I don't believe it is," she is contradicting herself.
For knowledge, the proposition believed must be true. A belief that is false cannot be considered knowledge because knowledge presupposes a relation between a person and some fact or state of the world. The truth condition ensures that this connection to reality is maintained. For instance, if a person believes that it is raining but it is actually sunny outside, she doesn't "know" that it is raining, because her belief does not correspond to reality.
Justification provides the rational grounding or support that connects a person's belief with the truth. For example, if someone randomly claims it is raining without any evidence (e.g., looking outside or checking the weather), and it turns out to be true, she doesn't know, even though their belief happens to align with reality.
Thus, the individual necessity of justification, truth, and belief is generally accepted. However, the real issue arises when we ask whether these conditions are jointly sufficient.
I will now focus on the sufficiency as this is a problematic part. I will argue that the sufficiency claim fails as there are cases where one believes that it is raining, one is justified in so believing and it is in fact raining, and yet one doesn't know that it is raining.
Following Gettier, who presented counterexamples to JTB, consider the following scenario:
Imagine Sarah is sitting inside her office with no windows. She needs to know whether it is raining outside, so she turns on the live feed from a reliable weather camera pointed at the street outside. Sarah watches the live feed and sees people walking with umbrellas, and the ground looks wet. Based on this, she forms the belief that it is raining outside.
Sarah has good reason to believe it is raining. The live camera feed is usually reliable, and she sees clear evidence of people using umbrellas and the ground being wet.
It is actually raining outside at that moment.
Sarah believes it is raining based on the evidence she observed on the live feed.
However, unknown to Sarah, the weather camera she is watching is not functioning properly. The footage she is viewing is not a live feed but a pre-recorded video from a rainy day a week ago.
Sarah's belief that it is raining is true, but it is true by coincidence. She happens to be correct that it is raining, but not because of the evidence she relied upon (the faulty live feed).
Her belief is justified because under normal circumstances, the camera feed would provide reliable evidence, and the use of umbrellas and wet streets reasonably suggested rain.
Despite having a true, justified belief, Sarah does not actually "know" that it is raining because her belief is based on misleading evidence (a malfunctioning camera). Her belief is true by luck, which means it does not meet the criteria for knowledge.
This counterexample highlights a fundamental issue with the JTB account, rooted in the nature of justification itself.
JTB assumes a fallibilist notion of justification. If justification were infallible, the truth condition would be redundant because infallible justification guarantees the truth of the belief. The very need for a truth condition reflects the fact that justification is often fallible—it does not always ensure that our beliefs align with reality. Therefore, the fallibilist nature of justification is essential to the JTB framework: it acknowledges that beliefs can be justified without being true, hence the need for both justification and truth as distinct conditions in the analysis of knowledge.
However, denying fallibilism has significant consequences. If we insist on infallibility for justification, we dramatically reduce the scope of what counts as knowledge. Human knowledge, by its very nature, is fallible—most of our justified beliefs do not guarantee truth, yet we still consider them to constitute knowledge. Scientific theories, historical facts, and even everyday empirical observations would fail to meet the standard of infallibility, leading to a form of skepticism where we could know very little. Thus, fallibilism is crucial because it allows us to retain a practical, realistic understanding of knowledge, one that fits with the way we actually form beliefs and acquire information in the world.
If fallibilism is true, justification does not need to be perfect or infallible; it can lead to mistakes. This means that someone can have a belief that is justified but not true. When a belief happens to be true by some unrelated factor, this disconnect between the justification and the truth becomes the basis for creating Gettier cases.
In constructing the counterexample I followed Zagzebsky's recipe. It suggests coming up with a false belief which meets the justification criteria and adding a good luck element which turns the false belief into truth. This recipe also works for the alternatives to justification criterion, such as warrant or reliability, etc.
This ability to generate counterexamples with ingenuity shows that the traditional JTB analysis is not sufficient for knowledge, since it allows for cases of true belief grounded in luck or coincidence rather than genuine knowledge.
Returning to our original claim, we can now see that it proves to be inadequate as a comprehensive account of knowledge. While we have established that justification, truth, and belief are individually necessary conditions for knowledge, Gettier-style counterexamples demonstrate that they are not jointly sufficient. The key insight lies in recognizing the fallibilist nature of justification, which allows for the possibility of justified true beliefs that do not constitute knowledge due to epistemic luck.