Spirits, Spandrels and Zombies

Abstract

The nature of subjective experience remains one of the deepest unsolved problems in science. Phenomena such as sentience, self-awareness, phenomenal consciousness, and the sensation of free will have occupied philosophers and scientists for centuries, yet they remain unexplained despite substantial progress in neuroscience, biology, and computer science. The central challenge of explaining how objective physical processes give rise to the irreducibly subjective character of experience remains as resistant to resolution as ever. In this work, I argue that resolving this challenge is not merely a philosophical obligation but a practical necessity for the construction of genuinely intelligent artificial systems, and I propose three principles to constrain and guide the search for an explanatorily adequate theory of consciousness. Principle 1 (Lovelace) holds that consciousness is the product of information processing and requires no non-physical substrate — therefore, no spirits. Principle 2 (Darwin) holds that consciousness has genuine causal effects on behavior and could not have been preserved by natural selection had it not—therefore, no spandrels. Principle 3 (Turing) holds that systems exhibiting identical behavior across all possible situations must possess internally equivalent representations, making behaviorally indistinguishable yet experientially empty systems impossible—therefore, no zombies. I develop each principle, examine their joint consequences for theories of consciousness and for the architecture of artificial general intelligence systems, and argue that accepting even one of their negations forecloses the possibility of a science of mind with genuine explanatory and constructive power.