The spark

1  The following is an essay I wrote last year. I found it quite interesting and fascinating how I even forgot that I wrote this essay. Hopefully, it will be a spark. 

     To what extent can knowledge be justified true belief?

“Justified true belief” or JTB, is currently the most widely accepted definition or approximation of knowledge. In my opinion, the three criterions of “justification”, “truth” and “belief” are essential to knowledge. For one to know, one must first believe in a proposition, that is true, and have means of showing that the proposition is true. For the purpose of discussion, the definitions of “justification” and “truth” will be pluralistic. Combining both evidentialistic and realibilistic viewpoints, “Justification” is defined as a means of proving a proposition by providing evidence or using reliable cognitive processes or faculties. “Truth” is defined as a correspondence of propositions to the external world as it is, a coherence of propositions to other truths and any judgments that produce satisfactory or pragmatic consequences. By considering criticisms to JTB, it will be concluded that JTB represents knowledge to a large extent and is our best approximation of knowledge yet.

It can be argued using Gettier problems that JTB is not restrictive enough. Consider the following example of a Gettier problem: A farmer observes a grassy hill from far and notices something white which appears to him as a sheep. He then recalls images of other sheep he has seen in the region and now believes that there is a sheep on the hill due to the strong resemblance between the object he sees and sheep he has previously seen. In reality, what the farmer actually saw was model of a sheep made by an artist to resemble a real sheep as closely as possible. However, behind this model lies a real sheep that could not be seen from the farmer’s position.

 In the example above, the farmer possesses “knowledge” according to the JTB criteria. Firstly, the farmer believes that there is a sheep on the hill. Secondly, this belief is true because it corresponds to the actual situation. Lastly, the farmer has his justification for believing, the sensory data he receives corresponds to images of sheep he has previously seen, hence, he has evidence and also used reliable cognitive processes to justify his belief. As a result, the farmer does possess knowledge defined as JTB even though he does not know that what he actually saw was not the real sheep. Such Gettier problems show that the JTB definition allows one to acquire knowledge through coincidence or luck. For many, this presents a flaw in the definition because the “knower” does not seem to have the appropriate understanding of why the belief is true. Hence, to rectify this flaw, it is proposed that an additional condition in addition to JTB is required.

One such attempt is to adjust the JTB definition so that the “justification” relied on no “false premises” in its chain of reasoning. In the Gettier problem introduced above, it can be argued that the farmer subconsciously makes the “false premise” that it is impossible that his sense perception is deceived by an object that resembles a sheep. Therefore, if a condition to disallow “false premises” was added, the farmer’s JTB will not be considered as knowledge. However, this adjustment to the JTB is also problematic. With the new condition, it becomes necessary “know” that a premise is true before using it in a justification. As a result, a premise used in a justification would be a JTB requiring another justification, leading to another JTB. Therefore, the attempt to disallow “false premises” makes the problem of justification regression more evident.

The problem of justification regression leads to another argument that the “justification” component of JTB is unnecessary as it is problematic and thus should be omitted, leaving with only TB as the two criterions of knowledge. However, this modification to JTB is also problematic. Firstly, it seems counter-intuitive that one does not have to provide any evidence or use any cognitive process to show that a belief is true. Secondly, it seems impossible for one to arrive at a conclusion that a belief is true without any justification. Lastly, it is only reasonable that “truth” is accompanied with “justification”.  The three theories of “truth”, correspondence theory, coherence theory and pragmatic theory define “truth” by considering how it should be tested. These tests of truth can only be made in conjunction with a justification. For example, in the correspondence theory, a judgment is true if it corresponds to how it actually is in reality. To test that a judgment is true, one must check for a correspondence, and this process of checking is a justification. Hence, the omission of “justification” from JTB is unproductive in terms of refining our approximation of knowledge.

In general, counter arguments against the JTB are arguments concerning the “justification” component of JTB. These arguments are valid in terms of showing a certain flaw in JTB. However, they do not lead us to a suitable modification of JTB. In my opinion, the current failure in further refining JTB can lead to two possible conclusions. Firstly, even with its limitations, JTB is the most appropriate approximation of knowledge yet. Secondly, the approach of defining knowledge as a set of independent conditions may be misguided. However, alternative approaches to defining knowledge have not given us equal insight as the JTB. Hence, my personal conclusion is that compare to other definitions, JTB represents knowledge to the largest extent.