i).Explain why the Gettier problem is a problem for fallibilism not infallibilism. ii). What must fallibilists do in order to solve the problem? iii). Do any fallibilists manage to offer a successful solution, or is the Gettier problem fatal to fallibilism?
In this paper I will support the claim that the Gettier problem is fatal to fallibilism by showing that there aren’t any compatible solutions from Clark and Nozick. Gettier cases are a problem for fallibilism as the knowledge produced can be false whilst still following JTB.
In order to understand what Gettier was refuting, we need to understand the principle of justified true belief.
S knows that P if and only if:
i) P is true (we cannot know a false claim),
ii) S believes that P and,
iii) S is justified in believing that P.
JTB was the accepted theory of knowledge until the Gettier paper was published.
Fallibilism is the idea that propositions can be known even if we are not completely certain with the justification. There is always the possibility of doubt held to the truth of the belief. It is not denying that we have knowledge, it is simply saying that things we take to be knowledge may turn out to be false in the future. However, we could have a justified belief that turns out to be mistaken, yet it is not said to be knowledge due to the principles of JTB.
Gettier put forward the idea that JTB cannot be sufficient for knowledge, rather propositions work out to be true accidentally. For example,
Tina puts her car keys on the table every night before she goes to bed. She believes that they will be there in the morning. Due to the fact she has put the keys there herself, she has no reason to believe otherwise. Therefore, she is justified in her belief. However, what Tina is unaware of is her son coming home from university early and therefore unexpectedly. He takes the car keys after she has gone to bed to see his friends. He gets back before she has woken up and puts the keys back where he found them.
Therefore, it is true that he keys are in the same place she left them when she wakes up.
However, even though she has a justified true belief, it is just coincidental that her belief was true. If her son had stayed out a little bit longer then the keys would not have been there when she woke. Gettier has shown that JTB is insufficient for knowledge, the 3 conditions of JTB do clearly not work on their own, bringing about the idea that a fourth condition must be produced.
Fallibilism is saying that you can know something yet there is still a high chance you are in the wrong. This doesn’t seem logical: to say you know something whilst admitting there is a chance of error – realistically then, you don’t really know it at all, and this is what Gettier was picking up on: that a fallibly justified true belief doesn’t result in knowledge. Therefore, fallibilism is a problem for Gettier as you cannot be certain on what you claim to know, there is always space to be wrong, and doing so, Gettier does not label such beliefs and propositions as knowledge. It is merely accepting a justified belief as true until proven otherwise.
Beliefs only being true due to luck is why Gettier says that JTB does not produce an account of knowledge as all three conditions have been satisfied yet no knowledge is gained. JTB is also not a good account of knowledge as it comes to a circular argument. It can be argued that justification requires knowledge and therefore the definition becomes circular. Within my case, it is clear that each of the 3 conditions of JTB are met, yet it seems wrong to say that Tina knows the keys are going to be there the next morning, simply because she put them there herself, there are a number of open possibilities that could have happened so that her belief was false. So, my case shows that JTB is not a good account of knowledge as it only accidentally turned out to be true, and could have easily been false, and therefore not knowledge.
A main aspect of the Gettier problem is that even justified true beliefs can be deemed lucky in a way inconsistent with knowledge. As above, in my Gettier case, Tina believes her keys will be there the next morning, is justified in believing her claim, and it is true. However, as mentioned, it was simply lucky that the keys weren’t moved, or her son didn’t arrive home any later. Therefore, she didn’t know her belief as it could have easily been different.
Therefore, the Gettier problem is a problem for fallibilism because it makes the wrong prediction about Gettier cases: that S knows that P. It can’t explain why S doesn’t know in G cases. In my Gettier case, fallibilism makes the wrong prediction that Tina knows her keys will be in the same place the next morning, because cannot be completely certain on this.
The Gettier problem is only a problem for fallibilists because a Gettier victim is nowhere near to having perfect justification, therefore infallibilists will deny they know. For example, within my Gettier case, Tina does not have perfect justification as in order to do this she must be 100% certain that her keys will be in the same place in the morning, yet there is no way she can keep track of her keys once she has gone to sleep, meaning her belief cannot be infallibly justified as it is an all or nothing approach to justification. Taking an infallibilist approach would enable knowledge to become limited and only consists of a priori truths.
Fallibilists have tried to combat the problem Gettier proposed in order to have a theory of knowledge that works. The intuitive answer to this is to simply add a fourth condition to the fallibilist theory of knowledge such as:
(4) Fallibile justification ensures that 1 and 3 aren’t true accidentally.
However, it is difficult to pinpoint what exactly counts as an accidental belief. Clark and Nozick both offer more complex solutions to the Gettier problem, as will be discussed below.
Clark says that in every Gettier case a proposition that is true, is believed based on a false background belief. So, in order to combat the idea that Gettier refutes, Clark offers a fourth condition to JTB:
(4) The belief that p isn’t based on a false belief.
However, this doesn’t work for most Gettier cases. Using my previous example, Tina’s belief that ‘The keys will be on the table when I wake up’ is based on the true perceptual belief: ‘I put the keys on there myself.’ The justified true belief is not based on any false belief but it is still not knowledge. So, Clark doesn’t seem to solve the Gettier problem, emphasising the idea that fallibilism will always fault in the path of Gettier.
Nozick’s conditions explain what a reliable belief is. His theory makes sure S’s belief tracks the truth of P. Analogously, a bank teller knows he is dealing with fake money if he believes that a bank note is fake when it is fake and genuine when it is genuine. Therefore, S knows P, if and only if:
It is true that P
S believes that P
If (in a nearby world) P was not true, S wouldn’t believe that P
If (in a nearby world) P was true, S would believe that P.
In order to have knowledge, all four conditions must be met. In my Gettier case, she believes that the keys will be in the same place they were when she went to sleep when she wakes up. This is true. However, it is only true when her son puts the keys back before she woke up. In a nearby world, it is true that she believes the keys will be in the same place when she wakes but this is only if the situation is slightly different and the keys are there when she wakes up (as p needs to be true). This certifies condition 4. When coming to condition 3, this is where the Gettier case falters. In the nearest world where it is false that the keys aren’t in the same place the next morning (because he didn’t get back before she woke up), she will still believe that proposition even when it is false. This is because she is still asleep and is therefore unaware her keys are gone. So, S doesn’t pass the third condition in the nearby world, meaning that she doesn’t know that her keys will be there when she wakes up in the actual world. It is not knowledge unless all four conditions are met. This is a common pattern in Gettier cases, they often fail condition 3 due to relying on justification from a while back. Nozick puts this forward as his way to show a connection between justification and truth, ruling out cases of luck. She is not sensitive to the truth of P (that her keys will be there when she wakes up) and this is why she doesn’t know. Therefore, Nozick does not solve Gettier as in a Gettier case S can hold a false belief when failing condition 3 and if they are holding a false belief then they don’t know the proposition.
Overall, fallibilism will always be caught out by the Gettier problem. This is because it doesn’t make sense to say that you have knowledge of something whilst allowing way for error, as if you are open to it being false, this means that you didn’t necessarily know it in the first place. Furthermore, solutions put forward from Clark and Nozick both fail in response to the Gettier problem and therefore a fallibilist cannot gain knowledge from a gettier case.