Doesn proof planning promote cheating by permitting ad hoc adjustments to enable a prover to `discover particular proofs?
Not if the recommended methodology is adopted. The science of reasoning paper specifies a set of criteria for assessing proof plans. These include generality and parsimony, which discourage the creation of ad hoc methods designed to guide particular theorems. Rather, they encourage the design of a few, general-purpose methods which guide a wide range of theorems. The expectancy criterion promotes the association of a method with an explanation of why it works. This discourages the design of methods which often succeed empirically, but for poorly understood reasons. Of course, these criteria are a matter of degree, so poor judgement may produce methods which other researchers regard as ad hoc. The criteria of proof planning then provide a basis for other researchers to criticise such poor judgement.
Related Questions
- How do you reflect budget adjustments on planning estimates for increases/decreases in federal funding when the amount of funds remains constant?
- How easy or difficult is it going to be to include short ad hoc courses and other elearning resources into the local catalogue?
- How does QD respond to ad hoc analytical read-only complex queries faster than a standard RDBMS on a PC?
