The basic purpose of the paper is simple; having proposed a set of axioms or 'basic truths' regarding Structural Health Monitoring (SHM) in a previous paper, the authors would like to extend the set by the proposal for a new axiom. This axiom relates to an observation that the presence of damage in a structure or system usually results in increased complexity of measured responses or features. It is argued that this observation could lead to principled means of selecting effective features for SHM.