Hard Problem of Consciousness

David Chalmers introduced the distinction in a 1994 talk in Tucson, in print as "Facing Up to the Problem of Consciousness" (1995). He split the study of consciousness into two problems:

• Easy problem: explain the brain processes that produce behavior, perception, attention, verbal report, and so on. Hard in the ordinary scientific sense, but in principle the kind of thing neuroscience and cognitive science can do.
• Hard problem: explain why those processes are accompanied by subjective experience at all. Even a complete account of the easy problem would leave an "explanatory gap" between brain function and the felt quality of experience.

The hard problem is usually phrased in three related ways: explaining qualia (the felt redness of red), explaining subjectivity (the bare fact that there is anyone "home"), and explaining what Thomas Nagel called the "what it is like" of an experience.

The cluster of supporting moves

Chalmers' position is reinforced by a set of thought experiments and related arguments:

• philosophical-zombies — a being physically identical to a human but without inner experience is supposedly conceivable, which Chalmers reads as evidence that experience is something extra on top of physics.
• The knowledge argument (Frank Jackson's Mary the colour scientist): a complete physical knowledge of colour vision doesn't tell you what it's like to see red.
• The inverted spectrum: two people could be functionally identical while one sees green where the other sees red.

These are framed as showing that the easy and hard problems are distinct.

Rovelli's objection

Carlo Rovelli (Noema, May 2026) argues the hard problem is a category error. It assumes upfront that there is a metaphysical gap between mind and body — the gap is not discovered, it is smuggled in by the framing. His positive view: scientific descriptions are always perspectival because knowledge is always embodied. The first-person and third-person accounts are not two realities; they are two perspectives on the same brain process. "Why does red look red?" is no more in need of explanation than "why does the cat look like a cat?"

This is a modernised version of Spinoza's monism: there is one nature, and mental phenomena are the same nature described in salient terms rather than complete ones. Rovelli's phrase: mental processes are physical processes by subtraction, not by addition.

Other dissolving moves

Rovelli is not alone. Variants of the same dissolution have come from:

• Daniel Dennett (consciousness as a set of competing processes; experience is "what it seems like" not "what it is")
• Patricia Churchland (eliminative materialism)
• Higher-order theories (consciousness as a representation of a representation)
• Predictive processing accounts (consciousness as the form of a generative model's self-model)

What Rovelli adds is a physicist's reading of why the gap-argument feels compelling even when it is wrong: people treat scientific pictures as views from outside reality, then notice that experience does not show up in such a view. The fix is to give up the view-from-nowhere conceit.

What this leaves untouched

Skeptics of the dissolution still owe an account of:

• Why neural activity is accompanied by experience at all (the bare existence question)
• How to verify, from the outside, whether a system has subjective experience (the LLM consciousness question, the animal consciousness question)
• What experience even is in machines that lack biological substrate

Rovelli's answer would be that the "bare existence" question is malformed — it assumes a category of "with vs without experience" that doesn't carve nature at the joints. But this is contested.

See also

• no-hard-problem-consciousness — Rovelli's full essay
• philosophical-zombies — the canonical Chalmers thought experiment and its self-defeating structure
• claude-emotion-concepts — interpretability research that finds causal emotion-related directions in LLM activation space; if Rovelli is right, this is the kind of work that dissolves the gap rather than confirms it