Pain still hurts. Grief is still heavy. The self still coheres. What dissolves is not the phenomena but the unnecessary category imposed on top of them. Dennett called consciousness an illusion. This paper argues he was right that there is an illusion — but wrong about what it is. The illusion is not experience. It is the false unity of the category.
Pain is real. Grief is real. Love is real. The self is real. The specific character of this moment — its quality, its texture, its weight — is real. Nothing that follows takes any of that away.
What the paper argues is narrower than it sounds: the word "consciousness," treated as the name of a single unified thing, does not pick out anything science has been able to find, define, or measure — after three centuries of trying. The phenomena are real. The category bundling them together is not.
This paper is more conservative about lived experience than illusionism, not less. It asks the reader to doubt a label, not their own pain.
The reader knows the phenomena are safe before the philosophical work begins. The constructive account is built before objections are addressed. Open questions are stated honestly.
What this paper argues, what it does not argue, and why the Moorean intuition ("I know I am experiencing something") is an ally of the account, not an objection to it.
Pain, grief, love, the self — each phenomenon examined and shown to survive the retirement of the category. The distance between the reader's position and this paper's conclusion is one word, not one world.
Descartes introduced the mind/body split to solve a theological problem. The theology was abandoned. The framework survived. "Consciousness" is the last structural remnant of Cartesian dualism.
Three centuries of failed definition. Four families of proposed definitions, each failing in a characteristic way. Dissociation evidence: the components of "consciousness" come apart routinely.
Five defensible primitives: input, internal state, state-change, self-modeling, output. Feelings as real internal states with causal power. Six levels of self-modeling depth. No sixth primitive needed.
Every phenomenon walked through. Category A: fully recovered with greater precision. Category B: reframed but intact. Category C: correctly absent — the Cartesian residue dissolves.
Animals, AI systems, and ethics. The conscious/not-conscious binary replaced by a spectrum of aversive-state capacity and self-modeling depth. Edge cases become tractable.
Seven objections in their strongest form. Qualia, illusionism, zombies, "Dennett said this," eliminativism, self-refutation, and the suffering threshold — each answered directly.
The threshold problem. A taxonomy of self-modeling depth. Integration with predictive processing. What cognitive science looks like after consciousness. What was dissolved — and what remains.
Dennett demystified consciousness. Frankish denied the phenomena. This paper retires the category while affirming everything it was pointing at. That is a structurally different move with different consequences.
The consciousness-based ethical framework is binary: conscious or not. The processing-system account replaces it with measurable dimensions — aversive-state capacity, self-modeling depth, state propagation — that handle edge cases by reflecting the biology honestly.
Replace it with: What is the complexity of its internal state space? Can it enter and sustain aversive states? Does it have self-modeling? How do its states propagate? These questions have graduated, empirical answers.
Current large AI systems have internal states, can refer to themselves, and produce outputs sensitive to their own processing. Whether this constitutes genuine aversive-state capacity or self-modeling is an open, tractable question — not an unanswerable metaphysical one.
Vegetative states, foetal development, animal welfare law — all currently framed around "consciousness," a concept nobody can define. Reframe around processing capacity, aversive states, and self-modeling depth. The questions become answerable.
"The real illusion was never experience. It was the false unity imposed on top of it." — The Real Illusion of Consciousness, 2026
Nine Parts plus a Coda. From reassurance through history, through three centuries of failed definition, through the constructive account and its recovery of every phenomenon, through the implications for animals and AI, through seven objections answered directly, to the open questions stated honestly. Discourse eliminativism: the phenomena survive; the label is retired.
"Pain still hurts. Grief is still heavy. The recognition of a face you love still produces something specific and powerful. The word was the problem, not the things it was trying to name." — The Real Illusion of Consciousness, 2026