Some Problems about Knowledge
Introduction
If I asked you to list all the things you know, it would at first be hard for you to know where to start. But once under way it would then be hard for you to know where to stop. The list is endless. Think of all the things that you know about your immediate surroundings. If you're sitting in a room while reading this, then you could list countless items of knowledge about the colours in the room, the position of the furniture, the contents of the room and so on. But not only do we have a seemingly endless supply of knowledge about our immediate surrounds but also about ourselves – our history, thoughts, memories, our friends and family, etc. We also have a huge store of of knowledge of a more abstract and general kind about history, science, mathematics and so on. For instance, we know that 1+1=2, 1+2=3, 1+3=4, etc. Our knowledge of arithmetic alone is enormous and perhaps infinite.Analysing knowledge
So we have lots of knowledge. But what exactly is knowledge? What is it to know something? How is knowing that p different from believing that p or being sure that p? Here's a first attempt at answering these questions. For you to know that p the following two conditions must be satisfied:1. You must believe that p.
2. P must be true.
So according to this analysis, knowledge is true belief. Thus, I know that Sydney is north of Melbourne because I believe that Sydney is north of Melbourne and it is true that Sydney is north of Melbourne. All's well it seems. But suppose that I'm sure that Aquanita is going to win the last race at Randwick and Aquanita does win the last race at Randwick. Then according to (1) and (2) I must know that Aquanita was going to win. Yet it is clear that I didn't know – there was a fair amount of luck involved. Conditions (1) and (2) are not sufficient. What else is required? Some sort of justification or basis is needed for any claim to know something. A third condition must be added:
3. Your belief that p must be rationally justified.
This condition excludes the Aquanita case as there is no real rational justification in horse racing (though there are a few rough guides such as form and the condition of the track). There is still a problem however. What makes a belief rationally justified? Suppose I have a true belief that p. What justifies this belief? It might be based on some other belief q. But if q is to serve as a justification for p then I must know that q. So q in turn must be justified by some belief r, and so on. There seems to be no point at which this sequence will stop. It can stop only if there is some sort of foundation for all our knowledge. Suppose that F is this foundation. Accordingly, all knowledge is somehow based on or rationally justified by F.
Leave the problem of finding a foundation for knowledge for the moment. The stage we have reached is that knowledge is justified true belief. Are (1), (2) and (3) adequate? Suppose that TV cameramen at the ABC have invented a very convincing form of trick camera work, which they keep a secret. Then on the 7pm news the ABC “shows” the Sydney Harbour Bridge collapsing. It's absolutely convincing, suppose, and fools everyone. So the viewers have a justified belief that the bridge has collapsed – they can see it for themselves on the ABC news (what more could you want?). Now suppose that while the bridge is being shown collapsing on TV it actually does collapse. The justified belief that the bridge has collapsed is now also true. Nevertheless, it doesn't seem right to say that the viewers know that the bridge has collapsed. For there is something wrong with the manner in which their belief has been justified. There is a causal element in knowledge. If I know that there is a book on the table there is a causal link between my knowledge and the book on the table. Similarly, if I know that Germany invaded Poland in 1939 there is a causal relation, albeit more complex and indirect, between that fact and my knowledge of the fact. So a fourth condition is required for the analysis of knowledge.
4. Your knowledge that p must be justified in the appropriate way.
This condition explains why the Harbour Bridge case is not an instance of knowledge. In this case the causal relation is between a television image and the viewers, but the bridge itself is not part of this causal chain. For it is being supposed that the TV image can be produced independently of the actual bridge. That's how the trick works. So the final account of knowledge is this: To know that p is to have true belief that p which is justified in the appropriate manner.
Some questions
- Do you think that the conditions (1) – (4) are sufficient for an account of knowledge? What more might be required?
- Does the account given require that knowledge have a basis or foundation? What could this foundation be? Perhaps: the knowledge of our mental states; or the knowledge of certain basic commonsense facts such as my knowledge that I am sitting in a room typing; or the knowledge of basic arithmetic. Does knowledge have a foundation, F say? How could we come to that F was the case? Our knowledge that F could not be based on anything else since F can't itself have a basis – it wouldn't then be a foundation for all knowledge.
- How can condition (4) acount for our knowledge of abstract truths such as those of mathematics? My knowledge that 2+3=5 does not appear to be causally related to the fact that 2+3=5. For what sort of fact would that be?
- Some of our knowledge is “non-propositional”. This means that some of our knowledge does not involve the truth of a certain proposition or statement. For example, I know what vegemite tastes like or what a harmonica sounds like but neither of these items of knowledge involve knowing the truth of some proposition or statement. Compare this with my knowledge that Sydney is north of Melbourne. This does involve knowing the truth of the statement “Sydney is north of Melbourne”. Do conditions (1) – (4) also account for non-propositional knowledge, or is a separate account needed for this?