evidence best explained by the supposition that the loaf of bread exists. Similar arguments support
inference to the existence of many other kinds of objects, from lions to mountains. Contrary to
empiricism, scientific knowledge does not come just from our senses, but goes beyond them via a
multitude of reliable instruments from telescopes and microscopes to Geiger counters (used to
measure radiation) and particle colliders (used to detect the behavior of subatomic particles). The
efficacy of scientific instruments is incompatible with idealism, because their measurements do not
depend on mental activity, but it fits well with constructive realism.
You might think that even if pieces of bread are real, their properties (color, taste, smell, and
texture) are not, because these are so heavily dependent on our minds. Many philosophers have
thought that nothing in the external world corresponds to people's experiences of colors, eliminating
them as real. Their arguments rely on the fact that there is no simple mapping between the space of
colors that people experience and the properties of objects that affect how they reflect light of
different wavelengths. Paul Churchland has found, however, a way of construing the physical
properties of objects that reveals a correspondence between their reflectance efficiencies and
people's experiences of colors like red, green, and blue. He describes how the human visual system
successfully tracks approximations of the reflectance profiles of objects at a low level of resolution,
so that colors can be viewed as objectively real properties of objects even if color vision is highly
context sensitive.
The correspondence between reflection properties and color experience makes sense given current
theories of how the brain processes color information, from stimulation of cells in the retina that code
for specific wavelengths of light to interpretations generated in the visual cortex. I like the conclusion
that colors are real properties of objects, and it does seem to fit with the best available understanding
of how the brain interacts with objects. But realism about objects could be true even if realism about
colors is not, as long as we have good reason to believe that objects and at least some of their
properties exist independently of mental representations of them.
I have tried to show in this section that the best explanation of the convergence of experiences from
the multiple senses of many people and instruments is that there really are physical objects that cause
these experiences. Moreover, the observable properties of these objects are much as we perceive
them to be. Of course, they have other nonobservable properties, such as their atomic structure, that
we can learn about only from scientific theorizing.
In sum, attention to how the brain functions in perception supports constructive realism over
empiricism and idealism. The constructive nature of perception with both top-down and bottom-up
processing shows the implausibility of a narrow empiricism that ties knowledge too closely to
sensory input. On the other hand, the robustness of sensory inputs of different kinds counts by
inference to the best explanation against the idealist view that the existence of objects is mind
dependent. Our perceptual knowledge is both constructed and about real things. Such constructive
realism is also the best approach to theoretical knowledge that uses concepts and hypotheses to go
well beyond perception.
Concepts
There is much more to knowledge of reality than sensory experience. Human discourse is full of
concepts, including knowledge and reality, that are not directly tied to what we can see, touch, taste,
smell, or hear. Philosophers, psychologists, and now neuroscientists attempt to figure out the nature of