Why Consciousness Is Hard

In 1995 David Chalmers made a distinction that reorganised the field, and its usefulness is partly in how often it is misread. The easy problems of consciousness are the questions of how the brain discriminates stimuli, integrates information, reports its own states, focuses attention and controls behaviour. The hard problem is why any of that is accompanied by experience.

He is explicit that the easy problems are easy only by comparison. Several will occupy science for a long time yet. They are easy in one specific sense: we know what an answer would look like. It would be a mechanism (this structure performs this computation, which produces this discrimination) and when we have the mechanism, the question is closed. For the hard problem we do not know what an answer would look like, and that is a difference in kind rather than degree.

The phenomenon the distinction points at has a name of its own. Joseph Levine called it the explanatory gap: even a complete physical account of a brain state seems to leave it open why that state feels as it does, rather than some other way, or none. Note that the formulation is epistemic. It is a gap in explanation, not an established gap in nature, and where you come down on whether it reflects something about the world is where you come down on the whole subject.

Chalmers sharpens it with the zombie. Imagine a being physically identical to you, atom for atom, behaving identically in every circumstance, with no inner experience whatsoever. If such a being is coherently conceivable, then the physical facts do not settle the facts about experience; something has been left out of the physical description. The standard reply attacks the inference from conceivability to possibility, pointing out that people have confidently conceived of things later shown impossible, and that our sense of what is conceivable is not a reliable guide to what could exist. That inference is the hinge, and almost everything else is downstream of it.

The strongest deflationary reply is worth stating properly, because it is usually caricatured as denying that anyone is conscious, which is not what it says. On this view we have two quite different ways of thinking about one thing. Physical concepts describe structure and function from the outside. Phenomenal concepts pick out states directly, from the inside, without describing them at all. Two such concepts can refer to the same state while never looking as though they do, and no amount of the first kind will ever resemble the second. If that is right, the gap is a fact about concepts, and nothing extra exists.

It is a respectable position and it leaves many people convinced that the question has been changed rather than answered. That reaction is itself worth examining rather than dismissing, because it is the same reaction people have to several successful reductions in the history of science, and it was wrong in most of those cases.

The practical use of the distinction is as a filter. When a team announces that it has explained consciousness by identifying how the brain integrates information globally, ask which problem has been addressed. Global integration is real and important, and theories built on it have done serious work on how states become available and reportable. But you can always ask the further question; why is the integrating accompanied by anything? And a functional theory cannot answer it, because functional theories are in the business of functions. That is not an objection to the work. It is a note about what the work is.