Sunday 29 April 2012

The Tripartite Definition of Knowledge

The definition of Knowledge as justified true belief, also known as the tripartite definition of knowledge has long and distinguished history and can be traced back to at least the Platonic dialogues. In Plato's dialogue Theaetetus, it is remarked that Theaetetus 'once heard that someone suggesting that true belief accompanied by a rational account is knowledge, whereas true belief unaccompanied by a rational account is distinct from knowledge'. (Plato 1987:201c-d). In this particular example the rational account consists in the justification for the true belief to count as knowledge. More recent adherents to Knowledge as Justified Belief include A.J Ayer who set out in his essay 'The Right to Be Sure' a version of the tripartite definition of knowledge. Ayer states that the sufficient conditions for knowledge as 'first that what one said to know is true, secondly that one be sure of it, and thirdly that one should have the right to be sure'(Ayer 2009:13). With justification just being synonymous for Ayer with being the right to be sure. You can see the difference between the formulation outlined by Ayer and Theaetetus but this distinction is purely how the tripartite definition of knowledge is phrased.

The Standard Tripartite Definition of Knowledge can be outlined as:  
s knows that p if: 
  1. P is true.
  2. S believes that p. 
  3. S has justification for his belief that p. 
There has been some discussion about whether justification and belief are necessary for knowledge. While it's possible to provide counter examples where it appears that a person can have knowledge without either one of justification or belief. The most notable example of such discussion is Colin Radford's example of a schoolboy who is one the receiving end of rapid string of questions by a teacher loses his belief in the truth of that p, but still answers the question correctly. This appears to be a case where belief is not necessary for knowledge, such examples are highly contentious. It also seems that the student provision of the correct answer is in some way indicative for a belief that p is true, otherwise any answer would be equally as conceivable. It was generally accepted that the standard tripartite definition of knowledge taken together provides sufficient conditions for knowledge ascription. 

This consensus was radically undermined when Edmund Gettier published a 3 page article called 'Is Justified True Belief Knowledge?' which appeared to destroy the standard tripartite definition of knowledge. Gettier provided two counter examples which appeared to undermine the tripartite definition of knowledge. He concluded that the two counter examples showed that the 'definition [of knowledge as justified true belief] does not state a sufficient condition for someone's knowing a given proposition.' (Gettier 2009:15) While the two examples originally provided by Gettier are valid they aren't particularly realistic. But it soon became apparent that it was possible to create numerous Gettier style counter examples and many valid counter examples have been formed. The classic example offered by Gettier is the case of Smith: 
Smith has applied for a job, but, it is claimed, has a justified belief that "Jones will get the job". He also has a justified belief that "Jones has 10 coins in his pocket". Smith therefore (justifiably) concludes  that "the man who will get the job has 10 coins in his pocket".In fact, Jones does not get the job. Instead, Smith does. However, as it happens, Smith (unknowingly and by sheer chance) also had 10 coins in his pocket. So his belief that "the man who will get the job has 10 coins in his pocket" was justified and true. But it does not appear to be knowledge.(Wikipedia Gettier problem)  

Numerous early responses to Gettier were touted to solve the problem with the traditional account of knowledge. Dretske's outlined a response that stated the reasons given in the justification of a knowledge claim must be the right ones, namely they must be strong enough to be conclusive. This is similar to the response made Chisholm which stated the evidence must be adequate. What is generally deemed wrong with these responses is that they set the bar of what counts as knowledge to high which in itself is problematic. It should be also noted that Gettier style counter examples can be produced that provide examples where the justification is based on stronger evidence. Again another related response outlined by both Lehrer and Swain set out what it can be called the indefeasibly reply. The situation must be that further information would not defeat the justification. Again this seems to set the bar for what counts as knowledge far to high, it's possible to think of plenty examples where some other conceivable piece of knowledge might defeat the justification.   

A number of alternative definitions of knowledge have also been offered, probably the most influential is Goldman's causal theory. For Goldman, S knows that p if and only if fact that p is causally connected in an appropriate way with S's believing that p. It seems that this approach seems to work in dealing with Gettier style counter examples (try it on a few examples). There are a still a couple of problems with the Causal theory as outlined by Goldman. One of the said problems is that is highly debatable that facts can stand in causal relationships arguably only events and agents can stand in causal relationships. Their are also other problems surrounding  the knowledge of universals and abstract mathematical objects (Goldman: says the causal account of knowledge only applies to empirical knowledge).  

By the far the most influential reply to the Gettier problem is Nozick's Conditional Theory which outlines Knowledge as truth tracking. Nozick's theory can be seen as an attempt to avoid problems with earlier causal theory and also an attempt to synthesize the best insights for earlier replies. Nozick's Conditional Theory:
  1. p is true
  2. S believes that p
  3. if p were true, S (using M) would believe that p
  4. if p weren't true, S (using method M) wouldn't believe that p
This seems to avoid all the problems faced by the other replies as well as providing a reply to Gettier cases. Saul Kripke offered a counter example to Nozick's second formulation of his conditional theory, but Nozick denies this counter example undermines his theory as he believes that Kripke's example fulfills all of the four criterion for knowledge. 

1 comment: