Action-at-a-distance is a generic property of physical theories. As such, it is not a fruitful idea in theory building. Its absence confers a rigidity on a theory which we exemplify through analysis of long-range correlations of the EPR-type in relativistic theories. Rigidity is desirable for fundamental theories, and theory building should focus on structurally unstable properties, making action-at-a-distance a side issue. Though apparent superluminal effects seem to be present in many present-day physical theories, we maintain that they are not a basis for action-at-a-distance.