Embodied realism can work for science in part because it rejects a strict subject-object dichotomy. Disembodied scientific realism creates an unbridgeable ontological chasm between “objects” which are “out there,” and subjectivity, which is “in here”. Once separation is made, there are then only two possible, and equally erroneous, conceptions of objectivity: Objectivity is either given by the things themselves” (the objects) or by the intersubjective structures of consciousness shared by all people (the subjects)

The first is erroneous because the subject-object split is a mistake; there are no objects-with-descriptions-and-categorizations existing in themselves. The second is erroneous because mere inter-subjectivity, if it is nothing more than social or communal agreement, leaves out our contact with the world. The alternative we propose, embodied realism, relies on the fact that we are coupled to the world through our embodied interactions. Our directly embodied concepts (e.g., basic level concepts, aspectual concepts, and spatial relations concepts) can reliably fit those embodied interactions and the underpinnings of the world that arise from them.

The problem with classical disembodied scientific realism is that it takes two intertwined and inseparable dimensions of all experience - the awareness of the experiencing organism and the stable entities of the structures it encounters - and erects them as separate and distinct entities called subjects and objects. What disembodied realism (what is sometimes called “metaphysical” or “external” realism) misses is that, as embodied, imaginative creatures we never were separated or divorced from reality in the first place. What has always made science possible, is our embodiment, not our transcendence of it, and our imagination, not our avoidance of it.