On Gettier and Klein: Amending our Traditional Account of Knowledge

In his paper, “Is Justified True Belief Knowledge,” Gettier refutes the traditional Justified True Belief account of knowledge by providing counterexamples that show that while the conditions provided by the JTB account are necessary, they fall short of being sufficient for knowledge. Klein’s paper, “A Proposed Definition of Propositional Knowledge,” suggests a fourth condition with which to amend the JTB account so that it provides both necessary and sufficient conditions for knowledge. In this paper I will explain the what JTB account of knowledge is and what it is designed to do, how and why Gettier’s counterexamples proves that this traditional account fails, and finally how Klein proposes to amend the JTB account of knowledge.

The traditional account of knowledge, known as the Justified True Belief account, states that in order to have knowledge, these three conditions must be true: first, the proposition P is true; second, the subject S believes that P; and third, S is justified in believing that P. In other words, S knows that P if and only if S has a justified, true, belief concerning P. The necessity of the first condition is obvious, since it can not be the case that we have knowledge of something that is false. It also seems obvious that we do not have knowledge of P if we do not believe that P. Can we claim to have knowledge of something we don’t believe? The last condition is the only one of the three that may not seem immediately obvious; however, if our knowledge wasn’t justified, then it would be nothing more than a lucky guess1.

Given the conditions specified by the JTB account of knowledge, the question to be asked is what exactly is this account designed to do? In providing an account, we aren’t concerned with the word ‘knowledge’, but with the concept that knowledge picks out. So, we aren’t trying to define what knowledge is, instead we are trying to give an account of what it is to have knowledge. As Steup puts it, when we are examining concepts such as knowledge and justification, we are interested in “what people have in common when they know something and when they are justified in believing something” (21).

If we are to give an analysis of the concept of knowledge, the analysans (that which does the analyzing) must specify the conditions that are individually necessary and jointly sufficient for the analysandum (that which needs to be analyzed). In laymen’s terms, for the JTB account to be correct, it must be the case that the conditions (i) P is true, (ii) S believes that P, and (iii) S is justified in believing that P, are each necessary and together sufficient for S to be qualified in knowing that P. In this (and in any) analysis, the analysandum, “S knows that P,” and the analysans, the three conditions, must entail each other and be necessarily coextensive. If this account is correct, if the three conditions are met, then it must be the case that S has knowledge, and vice-versa.

In Gettier’s famous paper, he provides counterexamples that show that the above analysis of knowledge fails. As we shall see, the Gettier counterexamples prove that while the conditions of the JTB account may be necessary, they are not sufficient for knowledge2.

One of Gettier’s counterexamples goes as follows: Smith and Jones have both applied for the same job. Smith has heard first-hand from their boss that Jones will get the job. Smith also knows (and don’t ask how) that Jones has ten coins in his pocket. Because of this information, Smith has evidence for believing the proposition: (a) Jones is the man who will get the job and Jones has ten coins in his pocket. The proposition (a) entails the following proposition: (b) the man who will get the job has ten coins in his pocket. Since Smith deduces (b) from (a), then he is justified in believing the proposition that the man who will get the job has ten coins in his pocket. However, it turns out that, unbeknownst to him, Smith also has ten coins in his pocket. Also, unbeknownst to Smith, it is he who will get the job, and not Jones. So, while proposition (b) is still true, proposition (a) has turned out to be false.

According to the JTB account of knowledge, Smith knows that (b) is true since all three of the conditions hold: (b) is true, since the man getting the job (Smith) has ten coins in his pocket; Smith believes that (b) is true; and Smith is justified in believing that (b), since he deduces (b) from (a), and was justified in believing that (a). But, is it really the case that Smith knows (b)? He believes (b), that the man who will get the job has ten coins in his pocket, but (b) is not made true by (a), that Jones is the man who will get the job and Jones has ten coins in his pocket. Instead, it is made true by the fact that Smith is getting the job and that he also has ten coins in his pocket. This counterexample shows that a person can be justified in believing a proposition that is deduced from a false proposition. Although Smith has a justified, true, belief, it doesn’t qualify as knowledge. Gettier has shown that the conditions of the JTB account can be met, yet one can still fall short of having knowledge. It seems that another condition must be added to our account of knowledge in order to circumvent such a problem.

We need an account of knowledge that neither includes nor excludes too much, and it seems as though our previous account includes too much, since it can be the case that the three conditions can be met, yet we can still say that S fails to have knowledge. We also need an account of knowledge that prevents a true belief from being a “lucky truth,” using Steup’s terminology. For a true belief to count as knowledge, it must be justified (it can’t be a lucky guess), but it also must not be a lucky truth, which is the case in the Smith and Jones example. Smith has a belief in a lucky truth; the belief is lucky because, due to certain facts (as opposed to certain evidence, such as is the case with lucky guesses), the truth wasn’t likely. Smith was justified in his belief, he was just lucky that he was right. So, whatever amendment we make to the JTB account, it must ban both lucky guesses and lucky truths.

Klein agrees with the above statement, arguing that a fourth condition is necessary to prevent lucky truths, or as he puts it, “felicitous coincidences”. The fourth condition Klein proposes is stated as follows: (iv) there is no such true proposition such that if it became evident to S at t, P would no longer be evident to S. To quote Klein, “If there is any true proposition D such that it and S’s evidence for P would make it unreasonable to expect that P is true, S doesn’t know that P” (62). In other words, there can’t be a defeater or a disqualifying proposition such that would make it the case that if S became aware of it, he would retract his knowledge claim.

How the Fourth Condition Ties in to the JTB account of Knowledge

In the case of the Smith and Jones example, the most obvious defeater would be the proposition, “Jones will not get the job.” If Smith knew that Jones was not going to get the job, then he would not deduce proposition (b) that, “the man who will get the job has ten coins in his pocket,” since it is unknown to Smith that he has ten coins in his pocket. Using Klein’s talk, if the proposition, “Jones will not get the job” became evident to Smith, (b) would no longer be evident to Smith. It is the case that Smith does not have knowledge of (b), which seems to be the correct conclusion. If Klein’s fourth condition makes it such that the analysans of the JTB account are jointly sufficient for knowledge, then it seems as though we have a correct analysis of knowledge.


1. We can define a lucky guess as follows: S’s belief in P is a lucky guess if and only if: (i) P is true; (ii) S believes that P; and (iii) S has no evidence (justification) for believing that P is true.

2. Gettier’s counterexamples are of a certain form: if S is justified in believing that P, and deduces Q from P, then S is just as justified in believing that Q as he is in believing that P.