Christof Koch has spent decades mapping the machinery of awareness, and he opened the 15th “Behind and Beyond the Brain” Symposium by pointing at the thing that machinery still can’t account for. A brain scan can show you exactly what happens when someone looks at the color red. It can’t tell you why seeing red feels like anything at all.
That gap has a name. Philosophers call it the “hard problem” of consciousness, and it’s the reason a field that keeps producing better maps keeps failing to produce an explanation.
What the scans can and can’t do
Neuroscience is good at localization now. Researchers can watch which regions light up when a person sees an object, feels pain, pulls up a memory or commits to a decision. Networks associated with awareness have been identified. None of that closes the distance between electrical and chemical activity and the fact that there’s something it’s like to be you.
Koch’s talk went after materialism directly, the dominant scientific position that everything that exists reduces eventually to matter, energy and physical law. He’s not dismissing it. He’s asking whether it’s finished.
Three places the model runs out
Koch plans to concentrate on three areas of uncertainty. One is whether conscious experience can be reduced entirely to physical processes in the brain. Another is what modern physics has done to the question of what counts as “real.” The third is the pile of human experiences that conventional models handle badly.
That last category includes near-death experiences, mystical states and terminal lucidity. Terminal lucidity is the strangest of the three: an unexpected return of mental clarity shortly before death, sometimes in people whose dementia or other conditions had already left them unable to communicate normally.
Koch isn’t claiming these prove consciousness can survive without a brain. He’s arguing they deserve serious investigation, because they may expose gaps in what science currently explains. Those are different claims, and the difference matters.
The philosophy he thinks is worth reopening
Two old ideas come back into play here. Idealism holds that consciousness or mind is central to the nature of reality. Panpsychism holds that consciousness isn’t restricted to humans or animals but may be a fundamental property distributed throughout nature in varying degrees.
Under either view, consciousness doesn’t switch on once a brain hits some complexity threshold. It’s already there, underneath. Koch thinks researchers may need to reconsider these frameworks while still applying modern scientific methods, which is a harder balancing act than it sounds.
The math that lets a thermostat into the conversation
Koch is a leading supporter of Integrated Information Theory, which tries to describe consciousness mathematically. The claim is that subjective experience arises in systems that combine large amounts of information into a unified whole.
The interesting part is the second variable. It’s not just how much information a system holds, but how deeply its parts are interconnected. A system with a high degree of integrated information could, on this account, have some form of subjective experience.
Which is why IIT has been described as a scientific version of panpsychism. It leaves the door open to consciousness existing in systems that aren't human brains, and it does so with equations rather than assertion.
The part with immediate clinical stakes
Koch is a researcher at the Allen Institute for Brain Science and previously served as a professor at MIT and Caltech. His work includes developing new approaches for detecting possible signs of awareness in patients who appear unresponsive.
Some people who can’t move or communicate may still have conscious brain activity. Reliable detection methods there aren’t philosophy. They feed directly into diagnosis, treatment and decisions about medical care, which is the sharpest argument for why the hard problem isn’t an academic indulgence.
The question Koch keeps circling is whether the brain creates consciousness or is connected to something more fundamental. He doesn’t answer it. What he does instead is refuse to pretend the current model already has.