A Simple Tricky Example

Suppose we have:

	s-result1 IF fact1 
     s-result1 IF fact3 
     ... 
     t-result1 IF fact1
     t-result1 IF fact4
     ... 

Now assume that the student gets s-result1 and t-result1, and the question is "Is fact2 consistent with those results?"

Since fact2 doesn't predict any results, it can't be ruled out directly by any result that does or doesn't occur. And yet fact2 is, in fact, provably inconsistent with the above results. Why? Because

Hence, even if the rules were constrained to single causes, which they weren't, you would still need something equivalent to a propositional theorem prover.

Back to the main discussion