Graves develops a general philosophical framework of "scientific realism," asserting that scientific theories have ontological import in determining the real structure of the world, and in particular that models play a central role as ontological hypotheses. He introduces two factors, the empiricist error and the logical errors that he feels have contributed to past misinterpretations of the nature of reality that were based on general relativity theory.
The historic continuities that join the metaphysics of Plato and Descartes with the monistic, organic nature of geometrodynamics are contrasted with the latter's radical break with the Newtonian or atomistic worldview which is predicated on a basic separation between matter and space. The Newtonian independence and distinctness of objects is at the polar extreme from their Einsteinian interdependence and continuity.
In discussing the physical principles of relativity, the author begins with a review of Einstein's original formulation ("classical" general relativity) and the significance of its basic assumptions and equations. He then presents the remarkable recent developments in geometrodynamics, which allow the program of identifying matter with space-time to be carried further than even Einstein suspected possible. The surprising discovery that electromagnetism can be incorporated into geometrodynamics without modifying Einstein's original equations appears to be formally correct, but reliance on multiply-connected topologies ("wormholes") to represent charge raises various unresolved questions.
Graves concludes that the present language of physics, like that of everyday life, is based on concepts of independence and separation, and that a wholly new language may be needed to describe the world in terms of geometrodynamics, in which space-time appears as the only substance, with curvatures as its attributes, and in which objects have no absolute individuality, distinctness, or location.