In his famous 1962 paper Philosophy and the scientific image of man, Wilfrid Sellars discusses the apparent conflict pointed out by Eddington between the manifest and the scientific images of a table. Whereas, according to the manifest image the table is continuously coloured all over, science tells us that it is composed of colourless molecules. Moreover, molecules are made of nucleons and electrons which lie far apart from each other. Sellars sides with Eddington in defending the primacy of the scientific image. However, Sellars contends that “the manifest and scientific images could merge without clash in the synoptic view”. Such synoptic view could be achieved in principle by a complete scientific explanation of sensations, of colour for example. 

While I agree with Sellars in granting some primacy to science, I will try to defend this view in a different way. Unlike Sellars, I contend that science is continuous with the manifest image. We are irremediably merged in the world of the manifest image, or “common sense”, as I prefer to call it, with no possible escape. Thus, there is only one image or framework we can epistemically have access to. I advocate a direct realism according to which we have immediate perceptual access to external properties of things and also a scientific realism according to which we have good reasons to believe in properties which are not directly observable but to which we have indirect epistemic access by relying on causal generalities or laws which are empirically justified by induction. As a consequence, beliefs in some “unobservable” entities are justified by observations and methods which are part and parcel of common sense. Contrary to Sellars, inference to the best explanation and explanationist strategies in general are, in my view, powerless to justify belief in “imperceptible” entities.

In the light of this, I'll propose a way of solving the two tables predicament. Statements such as “the table is brown” can be wrong, but only partially. Science can correct direct perception in the same way as we correct our perceptions in everyday life. For example, in the same way as science tells us that the tabletop isn't homogeneous, when we get closer to a green thing seen from a distance, we may come to realize that it is composed of green trees.