This paper argues that if conclusions about children's grasp of logical concepts are to be reached and acceptable lines of research followed, then more precise definitions of the concept "logical necessity" must be formulated. The paper defines logical necessity as "the unconditional guarantee of truth that accompanies valid deduction from axiomatic premises." It then shows that whereas hypothetical conservation and transitivity problems have solutions that are accompanied by logical necessity, physical embodiments of those problems as tasks do not. The remainder of the paper outlines problems associated with learning logical concepts from physical examples, and discusses the ambiguity of the word "must." It suggests a set of criteria for inferring a grasp of logical necessity and briefly discusses the consonance of the ideas presented with those of Jean Piaget. (FL)