The most important part of Koch’s Concsiousness: Confessions of a Romantic Reductionist comes near the end, when he suggests an experimental test of the so-called “Integrated Information Theory”. Per Koch, a conscious experience is determined by the complete state space of the system that describes it as well as the individual state currently occupied; thus, the experience of being a brain with a large population of neurons deactivated should should be qualitatively different from being a whole brain, even when those neurons are not actually being used.
There are numerous difficulties that arise here, but I want to address specifically the claim that the procedure suggested above—knock out an unused chunk of brain and ask the patient if anything has changed—is, as Koch claims, a test of IIT. The problem is that a negative result doesn’t disprove the theory. In IIT, as in every reasonable theory of mind and human biology we have, the contents of memory are propositional rather than subjective. If my subjective experience of the colors red and blue were interchanged, I would not be suddenly startled to find the sky a different color—the color would change in all my memories as well. So too for Koch’s test subject, who, with a piece of her brain frozen, would have no way to remember what it used to feel like, and would consequently report no change.
Indeed, a positive result would support claims much, much stronger than the ones Koch is making: either that qualia are somehow encoded directly in the grey matter (so that it’s possible to read off my experience of red or pain directly from the structure of my brain), or there exists in consciousness a kind of “memory” of past states which is independent of the memory that neurons record. So this is ultimately a test of a different theory (and indeed one with no proponents).