The+Tripartite+Theory+of+Knowledge

Source: Dupré, Ben. //50 Philosophy Ideas You Really Need To Know//. Quercus Publishing Plc., London, 2007. Pages 24-27.


 * 'Uh oh, wrong turn,' thought Don, as he saw the hated figure slumped against the lamp-post, the all-too-familiar features of his brutish face clearly visible in the yellow light. 'I should have known that that scum would surface here. Well, now I know... What are you waiting for, Eric? If you're so tough...' All his attention focused on the figure in front of him, Don didn't hear the footsteps approaching from behind. And he didn't feel a thing as Eric delivered the fatal blow to the back of his head... **

So did Don really know that his killer Eric was there in the alley that night? Don certainly believed that he was there, and his belief proved to be correct. And he had every reason to form such a belief: he had no idea that Eric had an identical twin called Alec, and he had a clear view of a man who was indistinguishable from Eric in every aspect.

**Plato's definition of kwnoledge.** Our intuition is that Don did not in fact know that Eric was present in the alley -in spite of the fact that Eric was indeed there, Don believed that he was there, and he was apparently perfectly justified in forming that belief. But saying this, we are running counter to one of the most hallowed definitions in the history of philosophy. In his dialogue Theaetetus Plato conducts a masterful investigation into the concept of knowledge. The conclusion he reaches is that knowledge is 'true belief with a logos' (i.e. with a 'rational account' of why the belief is true), or simply, 'justified true belief'. This so-called tripartite theory of knowledge can be expressed more formally as follows:

A person **//S//** knows proposition **//P//** if and only if:

1. //**P**// is true 2. //**S**// believes //**P**// 3. //**S**// is justified in believing **//P//.**

According to this definition, (1), (2) and (3) are the necessary and sufficient conditions for knowledge. Conditions (1) and (2) have generally been accepted without much debate -you cannot know a falsehood and you have to believe what you claim to know. And few have questioned the need for some form of appropriate justification, as stipulated by (3): if you believe that Noggin will win the Kentucky Derby as a result of sticking a pin in a list of runners and riders, you will not generally be held to have known it, even if Noggin happens to be first past the post. You just got lucky.

**Gettier's spanner in the works.** Much attention was predictably given to the precise form and degree of justification required by condition (3), but the basic framework provided by the tripartite theory was broadly accepted for nearly 2500 years. Then, in 1963, a spanner was very adeptly put in the works by the US philosopher Edmund Gettier. In a short paper Gettier provided counterexamples, similar in spirit to the tale of Don, Eric and Alec, in which someone formed a belief that was both true and justified -that is, met the three conditions stipulated by the tripartite theory- and yet apparently did not qualify as knowing what he thought he knew. The problem exposed by Gettier-type examples is that in these cases the justification for holding a belief is not connected in the right sort of way to the truth of that belief, so that its truth is more or less a matter of luck. Much energy has since been spent on trying to plug the gap exposed by Gettier. Some philosophers have questioned the whole project of attempting to define knowledge in terms of necessary and sufficient conditions. More often, though, attempts to solve the Gettier problem have involved finding an elusive 'fourth condition' that can be bolted onto the Platonic model. Many suggested improvements to the concept of justification have been 'externalist' in nature, focusing on factors that lie outside the psychological states of the putative knower. For instance, a causal theory insists that the promotion of true belief to knowledge depends on the belief being caused by relevant external factors. It is because Don's belief is causally related to the wrong person -Alec, not Eric- that it does not count as knowledge. Since Gettier's paper, the search for a 'patch' has developed into a sort of philosophical arms race. Attempted enhancements of the tripartite definition have been met with a barrage of counterexamples intended to show that some flaw is still there. Proposals that apparently avoid the Gettier problem tend to do so at the cost of excluding much of what we intuitively count as knowledge.

One suggestion for the fourth condition to supplement the tripartite theory is that knowledge should be what philosophers call 'indefeasible'. The idea is that there should be nothing that someone might have known that would have overridden the reasons they had for believing something. So, for instance, if Don had known that Eric had an identical twin brother, he would not have been justified in believing that the man leaning against the lamp-post was Eric. But by the same reasoning, if knowledge has to indefeasible in this way, Don wouldn't have known that it was Eric even if it had been. This is the case whether or not Don knew of twin brother's existence; there could always be a sense in which knowers never know that they know. Like many other responses to the Gettier problem, the demand for indefeasibility risks setting the bar so very high that little of what we usually count as knowledge will pass the test.
 * Should knowledge be indefeasible? **

[...]