In order for Justified True Belief to be sufficient criteria to be able to claim definitely that knowledge exist, there could be no exception to JTB being knowledge. However, in 1963, Edmund L. Gettier presented two cases which have been generally accepted by philosophers to be cases in which JTB existed but knowledge was not present, thus proving JTB insufficient. I, however, take issue with the significance of Gettier's argument and will attempt to demonstrate that it ultimately show nothing important about the JTB definition of knowledge.
Gettier's Second Scenario
I will first address Gettier's second scenario, since it is the more transparently flawed. In this scenario, the character of Smith believes that Jones owns a Ford on the justification that he's seen Jones driving a Ford every day. If this were all that was involved in the situation the JTB criteria would not be met because, in fact, Jones only rents a Ford. Thus, while justified, Smith's belief would in fact not be true and thus fail as knowledge according to JTB. Gettier further tweaks this scenario, however, to create a situation where all the criteria for JTB are met but Smith still does not possess knowledge. Smith, Gettier explains, does not know where his old friend Brown lives and holds the belief that either Jones owns a ford, or Brown lives in Boston. Unbeknownst to Smith, Brown actually does live in Boston. Since, as was already established, Jones does not own a Ford the belief is actually true on all counts and, according to Gettier, justified on the basis that Smith has seen Jones driving a Ford.
The flaw in this scenario is that Smith's belief is not at all justified. It is true that Smith is justified in believing that Jones owns a Ford (which is in fact false, and thus not knowledge) but he has absolutely no justification for the clause that states "or Brown lives in Boston." First, he has no justification for believing that Brown lives in Boston, he has absolutely no idea where Brown lives at all. Second, he also has no justification for creating a dichotomy between Jones's ownership of a car and Brown's location. For a belief to be justified all parts of it must be justified. For example, if I were to claim the following:
a. I am sitting in a chair
b. All people who sit in chairs are aliens
c. I am an alien because I am sitting in a chair.
I would clearly not have a JTB. While it is in fact true that I am sitting in a chair, it is false that all people who sit in chairs are aliens and thus my belief, while justified in part, is not knowledge because it is not fully justified. Gettier's scenario could be written in a way that demonstrates this.
a. Smith sees Jones driving a Ford regularly
b. Smith doesn't know where Brown is
c. Smith concludes that either Jones owns a Ford or Brown lives in Boston
Cleary, the belief in this scenario is in no way justified simply because a small portion of it is. Thus, Gettier's second scenario fails to prove anything about JTB.
Gettier's First Scenario
Gettier's first scenario appears at first to be more solid than his second one because the character involved does in fact seem to possess JTB. The problem involved here is one of the functioning of the human mind, which is entirely relevant to the question of when knowledge exists because the mind is the place where knowledge exists.
In Gettier's example, Smith and Jones are both applying for a job and Smith is told by the president of the company that Jones is going to get the job. Thus he believes that Jones is going to get the job on the justification of the company presidents word. This belief is, however, untrue as Smith actually gets the job. As in the other problem Gettier does not leave the issue at this but complicates things. It turns out that both Smith and Jones have ten coins in their pockets. Smith does not know how many coins he has in his pockets, but he has counted the coins in Jones's pockets and knows the number of them for a fact. Being justified in believing that Jones will get the job, Smith concludes the following "The man who will get the job has ten coins in his pocket". Of course, Smith gets the job and has ten coins in his pocket and thus that statement is true. However, because of the lack of link between the actual truth that verifies the belief and the justification for the belief demonstrates, according to Gettier, that this JTB is not knowledge.
Gettier's scenario is, however, missing an important dimension about how thoughts occur in the mind that is relevant to the problem. All thoughts and beliefs that occur in the mind have associations, both long term and short term. Thus if I think of the word "Father" it might occur in association with feeling and images related to my own father. This can be demonstrated on paper as "Father(mine)". If however I was at my friends house and thought or said something in reference to his father the association would instead be with feelings and images related to his father and could be written as "Father(friend's)". Thus, while on paper the two instances of the world would traditionally appear identical, the word existed in my mind as two separate things because of the associations. These associations are impossible to avoid and are the very reason we have a distinction between the denotation of a word and its conotation.
In the case of the Gettier problem, I would contend, when Smith thinks to himself (believes) that "The man who will get the job has ten coins in his pocket" he is associating "man" with Jones because the statement is made on the justification that Jones has ten coins in his pocket and will apparently get the job. Thus what Gettier made to appear one statement is actually two. The first is "The man(Jones) who has ten coins in his pocket will get the job" and the second "The man(Smith) who has ten coins in his pocklet will get the job". The first belief is entirelly justified but is in fact not true and thus fails as knowledge according the JTB criteria. The second statement, on the other hand, is neither justified nor held as a belief but is in fact true and thus fails as JTB. In other words, when Smith thought that "The man who will get the job has ten coins in his pocket" he was referring to Jones and was thus wrong. It not possible for a thought based on a justification to appear without the referent justification being associated to in the mind. Because there are in fact two separate statements which only appear on paper to be identical, the apparent flaw in JTB that Gettier attempted to demonstrate does not actually exist.
To further demonstrate my point let me create another scenario similar to the one that Gettier made. Smith and Jones are both still applying for the same position and still have ten coins in their pockets and Smith is aware of the money in Jones's pocket but not his own. Smith is also still told that Jones will get the job and concludes "The man who has ten coins in his pocket will get the job." The end of the story is different though, the company looks at their finances and decides they can't afford another employee and they close the position. Neither of the men get the job. Across the country in Boston, however, there is a man named Brown applying for a different job. It turns out that Brown also has ten coins in his pocket and he gets the job. Thus, the man who has ten coins in his pocket does in fact get the job. Except, of course, "the job" being referred to is a different job. The number of possible situations that a stated belief could possibly reference must be limited when the belief is being tested for knowledge.
Essentially, you run into the problem of when something becomes a heap (One grain of sand is not a heap, but how many make a heap? Where is the cut off?). Where do you cut off a scenario as relevant to the belief at hand? Do you say that only Smith and Jones job and ten coins apply? What about the bum down the street trying to get someone to let him wash their windows who also happens to have ten coins in his pocket? Brown in Boston? A little boy with ten coins in his pocket asking his neighbor to let him mow their lawn for money? If any possible situation that could be referenced by a belief is included then no criteria could ever be good enough to determine if a belief was knowledge. Ultimately, for any criteria for determining whether or not a belief is knowledge to be useful, only the situation that the belief is actually applying to can be tested.
Note: Gettier's actual article can be found at http://www.ditext.com/gett
