CRITERIA OF CONSCIOUSNESS: DIFFERENTIATED INTEGRATION OF BIOELECTRIC ACTIVITY AS A MEASURABLE MARKER

This paper proposes an operational approach to consciousness that addresses not the question of subjective experience but the practical one: which systems have consciousness and how to measure it. Relative to the previous version, a substantial error has been corrected. The defining marker was previously taken to be the synchronization of bioelectric activity: the higher the synchronization, the more consciousness. This contradicts the very data the paper relies on. The Perturbational Complexity Index takes low values precisely for globally synchronized responses characteristic of deep sleep, anesthesia and coma; during an epileptic seizure synchronization is maximal while consciousness is lost. Integrated information theory requires integration and differentiation together, and uniformly synchronous activity yields low Φ. In the present version the defining marker is the differentiated integration of bioelectric activity — the combination of connectedness and diversity that Φ and PCI actually measure. The correction gives the criterion explanatory power: it now explains why deep sleep, anesthesia and seizure are unconscious, which the previous formulation could not. The definition and the empirical claim have been separated. Operationally, consciousness is defined through the capacity of a system to sense, adaptively reshape its response and move purposefully as a whole; the adaptivity requirement excludes devices with a fixed input-output characteristic. The claim that this capacity is realized by differentiated integrated bioelectric activity is made empirical and therefore refutable. The contested status of integrated information theory is noted; Φ and PCI are used as empirical measures independently of IIT's philosophical claims. The membrane potential range is corrected: −60 to −90 mV applies to neurons, while the typical bacterial value is about −150 mV. It is separately stated that the approach deliberately brackets the "hard problem" of subjective experience, and that extending integration measures to neuron-free systems remains an unfulfilled part of the programme.