The Gettier Problem
Why is the ‘justified true belief’ definition of knowledge inadequate?
Knowledge as Justified True Belief
The classical account of knowledge claims that knowledge can be defined as a justified, true belief (JTB). This is also known as the traditional account or the tripartite account of knowledge.
It would be helpful to read the blog What counts as ‘knowledge’? for you to grasp this concept better.
This definition of knowledge as JTB was widely accepted until philosopher Edmund Gettier pointed out its deficiencies in his short but devastating paper, which will be discussed in short below.
I will be using Duncan Pritchard’s excellently written book — What is This Thing Called Knowledge? as a reference for this blog.
An Overview of Gettier’s Paper — Is Justified True Belief Knowledge?
Edmund Gettier, in his seminal paper “Is Justified True Belief Knowledge?” (p. 1963) challenges the traditional definition of knowledge as ‘justified, true belief’. He provides examples of Gettier cases which exemplify how these three conditions are not sufficient to qualify something as knowledge.
Through his two counterexamples, Gettier aimed to explain how there can be cases in which a person has a true, justified belief, and yet it fails to constitute knowledge.
Gettier Cases
From Gettier’s Paper: Suppose Smith and Jones have applied for the same job. The president of the company has informed Smith that Jones will get the job. Smith had also counted the number of coins in Jones’ pocket (which happened to be 10). Thus, Smith would be justified in forming the belief — The person who will get the job has 10 coins in their pocket. But unbeknownst to Smith, Jones will not get the job. Smith will. Smith is also unaware that there are 10 coins in Smith’s own pocket. Thus, Smith’s justified belief (The person who will get the job has 10 coins in their pocket) only happens to be true. The basis for this belief (the number of coins in Jones’ pocket) is not what makes it true. Instead, it is the number of coins in Smith’s own pocket (that Smith is unaware of) that does.
From Duncan Pritchard’s book: Suppose one morning John sees the time 8:20 on the clock in his hall. Thus, he comes to form the belief that it is 8:20 am. John is justified in forming this belief as this is usually the time he enters the hall in the morning, and the clock has been reliable for years, for accurately telling John the time. He has no reason to doubt the accuracy of the clock. But suppose the clock had stopped working exactly 24 (or 12) hours ago. Even though the time at the moment might actually have been 8:20, John would still have formed his justified true belief without having what one can call ‘knowledge’ of what time it is. That these two (time John sees on the stopped clock and the actual time) coincide is merely a matter of luck. Had John entered the hall even a minute before/after, he would have formed a false, justified belief.
These two cases show that knowledge is not merely justified, true belief.
How to construct Gettier cases
All Gettier cases have a common form. One thing that is crucial for a case to be a Gettier case is the presence of a justified true belief. The three steps for constructing a Gettier case are:
- Consider a person who forms their belief in such a manner that they would usually be forming a false belief (in reference to Pritchard’s example — a person looking at a stopped clock to figure out what time it is).
- Add a point to support the accordance of justification to the person’s belief (in continuation of the example above — that the clock is known to be reliable).
- Construct the case in a manner that in instances other than this one, the person would have formed a justified false belief, but because of mere chance/luck, happens to form a justified true belief (like when a stopped clock happens to tell the right time twice a day).
Responses
There has not yet been any decisive response to the Gettier problem. The initial thought that the traditional definition of knowledge could simply be altered a bit to escape this problem soon proved to be lacking.
One could say that in order to have knowledge, a person’s true, justified belief must not be based on any false presupposition. But there is difficulty in understanding and defining what ‘presupposition’ actually means. It needs to be strong enough to eliminate Gettier cases but not so strong as to prevent all/most knowledge acquisition.
Take Duncan’s example of John and the stopped clock. Suppose his sister Sally has a belief that the clock in her own hall is maintained regularly, even though it is not. But Sally does come to be aware of what time it is when she looks at the clock. If our understanding of the concept of ‘presupposition’ is liberal, then her false presupposition would restrict her claim at having knowledge of the time, even though looking at a clock is considered a good way to acquire such knowledge. But if we try to adopt a more restrictive view and do not think of Sally’s belief as a presupposition, how can we ascribe John with having a presupposition that he is right about what time it is because he is looking at a working clock (even though it is not a working clock)? Sally and John both have false beliefs about the clock. Shouldn’t both of them be seen as lacking knowledge of what time it is, then?
But what about Gettier cases that do not involve presuppositions at all? Take the case of a farmer who forms a belief that there is a sheep in the field, when she is really looking at a shaggy dog that looks very similar to a sheep. It so happens that right behind that dog is an actual sheep. The farmer’s justified, true belief is immediately formed, without any presupposition, on immediate perception of the shaggy, sheep-like dog. It does not seem like she has knowledge of the sheep being there in the field, though, since she did not, herself, perceive the sheep.
On the way to developing a Gettier-proof definition of knowledge
Epistemologists came to realize that merely tweaking the traditional account of knowledge to evade the Gettier problem was an impossible task. A radically new analysis of knowledge, such that it is Gettier-proof, is required.
When it comes to a proposition being regarded as knowledge, the traditional account of knowledge imposes one condition on the world — that it be true (and reflect reality), and two conditions upon the agent — that they believe, and be justified in their belief of the proposition. The latter is not dependent on the world and how things actually are, as one could, for example, form a belief based on a hallucination. What Gettier cases make us realize is that more should be asked of the world than just truth for one to have knowledge. Such a requirement that posits the need for the agent to form their belief in the right kind of way needs (apart from the belief just being true) to be figured out.
Food for thought — The Problem of Criterion
The problem of criterion refers to the epistemological struggle of figuring out what conditions must be met by something for it to count as knowledge. The Gettier problem has shown us that the JTB criteria is not satisfactory or complete in itself. Something needs to be added to it, but philosophers are not sure as to what exactly that is.
You can read The dilemma of defining ‘knowledge’ — The Problem of Criterion to better understand this problem.
The criteria of knowledge does not seem to be obvious or a-priori. So how is it that we are able to identify cases of knowledge without first understanding what its criteria or definition is?
