Light and magnetic and electrical phenomena were assumed to be explainable in mechanical terms. Representation of values possessed in terms of the states of the objects in question in classical physics is straightforward and intuitive: a state is just a ‘list’ of the quantities an object has, and the ‘degree’ which they are possessed, that evolves in accordance with an equation of motion. Dispersion is an attribute of a probability distribution, or more generally of a measure, and it indicates the way in which values spread on either side of the mean or ‘expectation value’. Relativisation to the preferred determinate observable gives rise to a class of n-tuples of relations, which people can think of as intantiations of ‘lawlike’ connections that always involve the preferred observable. Two alternative interpretations of quantum mechanics, the standard and the modal interpretation, were used to support this viewpoint.