Why the AI Consciousness Debate Does Not Resolve: The Recognition Question Is Not the Constitution Question

The paper argues that consciousness has never been medically defined, only defined legally and socially, with definitions varying across jurisdictions and centuries. Since there is no diagnostic biomarker for phenomenal experience, observers have historically judged other humans to be conscious based on observer-side grounds rather than constitutive facts, as shown by slavery, colonial classification, dehumanization, and shifting clinical categories. The AI consciousness debate conflates two distinct questions: the constitution question (whether a system meets a theory's conditions for consciousness) and the recognition question (whether an observer judges a system to be conscious, determined by the observer's beliefs and background).

Study at a glance

Abstract

Consciousness has not been medically defined. It has been defined legally and socially, and those definitions have varied across jurisdictions and across centuries. The DSM does not define consciousness. The ICD does not define consciousness. There is no diagnostic biomarker for phenomenal experience. The AI consciousness debate has been proceeding as if a constitutive specification of consciousness were available against which systems could be checked. The historical record shows what recognition has looked like in the absence of such a specification. Observers have judged other humans to be conscious or not on observer-side grounds, with the same constitutive facts present in observer and observed, in patterns that cannot be explained by constitutive differences. Slavery, colonial classification, the dehumanization literature, the shifting clinical and developmental categories, and the case of behaviorist scientists denying the scientific standing of consciousness while themselves being conscious are independent records of the same pattern. If recognition of consciousness in beings constitutively identical to the recognizer has been observer-dependent for the entire historical record, there is no observer-side basis to expect recognition of consciousness in AI to be governed by AI's constitutive facts, regardless of what those facts turn out to be. The paper diagnoses the AI consciousness debate's failure to resolve as a conflation of two distinct questions. The constitution question asks whether a system instantiates the conditions a theory of consciousness specifies; it is conditional on a theory, and no theory has been medically settled. The recognition question asks whether an observer judges a system to be conscious; it is a fact about the observer, determined by the observer's prior beliefs, training, and social background (Stewart, 2026w, *Cognitive bias as default setting*). The historical record refutes the assumption that the two questions can be answered together by arguing the constitution question alone. The paper supplies the separation, identifies what each question requires to be answered, and shows that the recognition question has been the operative question by default because no medical specification of consciousness has been available to anchor the constitution question. The paper does not advocate a position on whether any current AI system is conscious. The deferral is a deliberate restriction on what the paper claims, not an avoidance. **Keywords:** artificial intelligence, consciousness, perception, anthropomorphism, disorders of consciousness, philosophy of mind, AI ethics, machine consciousness