Speech rhythm has been investigated from phonetic, phonological and signal processing perspectives, leading to a wide range of non-comparable methodologies. We assume that this is because speech rhythm is an emergent phenomenon (Emergent Rhythm Theory), due to many ‘hidden' physiological and other factors, and that specialists make different selections from these factors. We also claim that models proposed so far are relatively arbitrary, and that their formal and empirical similarities and differences are not clarified. Although we accept Emergent Rhythm Theory, we acknowledge that it is currently too complex and inexplicit to be falsifiable, and concentrate on more highly constrained Physical Rhythm Theory approaches. We propose a set of explicanda for Physical Rhythm Theories, and formally and empirically compare selected single-parameter physical rhythm measures. We find that the selected measures show a very low degree of similarity and outline the steps necessary to improve the situation.