But, the argument goes, if natural and synthetic kinds are indefinite at the edges, and if the structures or forms of things, measured precisely enough are not the same as the ideal forms, then we do not really know the forms or natures of the individual cases and have to get along with conventions about them.