The cohomology table of a pair is defined as the family of non-negative integers . We say that a subclass of is of finite cohomology if the set is finite. A set is said to bound cohomology, if for each family of non-negative integers, the class is of finite cohomology. Our main result says that this is the case if and only if contains a quasi diagonal, that is a set of the form {(i,ni)i=0,…,d−1} with integers n0>n1>>nd−1.
We draw a number of conclusions of this boundedness criterion.