It is perfectly possible to prove some negatives, as WG has pointed out -- as I said further up it depends on the nature of the proposition being "proved". But in general, in engineering circles, the principle is to form requirements in positive terms to make testable propositions out of them.

A