General anesthetics are remarkable small molecules that transiently erase consciousness. They eliminate the sense of “me” together with sensory perception, movement, and memory, while leaving complex autonomic and housekeeping functions, and even early-stage sensory processing, mostly unchanged. Anesthetics also slash information content by synchronizing burst-firing in large populations of cortical neurons (slow-wave electroencephalogram [EEG]), and lessening functional connectivity (Pal et al., Reference Pal, Li, Dean, Brito, Liu, Fryzel and Mashour2020; Volgushev, Chauvette, & Timofeev, Reference Volgushev, Chauvette and Timofeev2011). Both changes should sharply reduce Φ, the integrated information theory's (IIT) metric of consciousness. Two contrasting models attempt to explain how anesthetics work: direct suppression of spinal and cortical activity, and interference with dedicated circuitry-designed evolutionarily to switch between consciousness and unconsciousness (Minert, Yatziv, & Devor, Reference Minert, Yatziv and Devor2017). Both models meet IIT's requirement that reduced Φ should reduce consciousness. Sometimes, however, when cortical Φ is expected to collapse, no equivalent loss-of-consciousness is observed. Some examples:

• • Unilateral intracarotid delivery of anesthetics (Wada test) induces slow-wave cortical EEG bilaterally. This should drastically reduce Φ, but subjects retain consciousness and memory formation (Halder, Juel, Nilsen, Raghavan, & Storm, Reference Halder, Juel, Nilsen, Raghavan and Storm2021).

• • Cortical slow-wave EEG also dominates non-rapid eye movement (NREM) sleep, presumably with low Φ. Yet NREM sleep supports dreaming, including the subjective sense of “being there,” just like in REM sleep with its wake-like EEG (McNamara et al., Reference McNamara, Johnson, McLaren, Harris, Beauharnais and Auerbach2010).

• • IIT sees the thalamo–cortical system as the maximally differentiated/integrated “main-complex” that generates high Φ (Tononi & Sporns, Reference Tononi and Sporns2003). However, large thalamic lesions in animals and humans can be compatible with retained or recovered consciousness (Fuller, Sherman, Pedersen, Saper, & Lu, Reference Fuller, Sherman, Pedersen, Saper and Lu2011; Jimenez Caballero, Reference Jimenez Caballero2010; Sakamoto, Okubo, Kanamaru, Suzuki, & Kimura, Reference Sakamoto, Okubo, Kanamaru, Suzuki and Kimura2015).

• • Large occipital cortex lesions surely reduce Φ. But the result is perceptual blindness, not impaired consciousness. It's like turning off the lights. The unawareness of visual stimuli is like the unawareness we all have of magnetism and ultraviolet stimuli. A skill is lost, but “I” am still present. Similarly, surgical hemispherectomy, prefrontal lobotomy, and congenital cortical thinning may cause disability, but not necessarily disordered consciousness (Feuillet, Dufour, & Pelletier, Reference Feuillet, Dufour and Pelletier2007).

• • Auras in epilepsy, and direct cortical stimulation, are almost never painful, and cortical lesions rarely if ever cause analgesia (Penfield, Reference Penfield1975). However, broad consensus sees consciousness as essential for experiencing pain. It follows that cortex is not necessary for experiencing pain or being conscious.

• • Conversely, stimulating certain subcortical sites does evoke pain reliably, and small brainstem lesions frequently induce unconsciousness, including loss of pain. Simple logic demands that the brainstem, perhaps its super-integrative isodendritic core, should be a lead candidate as “main-complex” and generator of high Φ and consciousness. But, in practice, proposals that the “primitive” brainstem might be primary are rarely taken seriously and are typically dismissed with metaphors like “the brainstem is a switch…” or “just a power-supply….” Ignoring the fact that mitochondria supply each neuron's power, should feelings of indignity drive scientific theory? Consider the following realignment of the roles of cortex versus brainstem in the generation of consciousness.

1. (1) The brainstem did appear early in evolution. But, with so little modern research dedicated to it, are we certain it stagnated ever since? In all vertebrates including man, the brainstem exerts broad control over the entire neuraxis including active control of state-of-consciousness. And its unique architecture, particularly the isodendritic structure of neurons in its reticular core, appears specialized for multimodal integration (Hobson & Scheibel, Reference Hobson and Scheibel1980; Ramon-Moliner & Nauta, Reference Ramon-Moliner and Nauta1966). Can one confidently reject the possibility that the raw feel of being conscious resides here, not in the cortex? Note also the consequences of this question for vegetative patients with an intact brainstem who appear to be awake but are currently presumed to be unconscious and incapable of experiencing pain.

2. (2) Over time, species evolved ever improving faculties to optimize niche-adapted engineering skills in the realms of sensory processing, motor performance, memory, and cognitive/predictive abilities. This process accelerated in early mammals with the appearance of cortical brain architecture. The “cortical” arrangement of neurons, in contrast to the “reticular” and “nuclear,” is optimized for fast, computationally efficient, and presumably unconscious information processing (Kaiser, Reference Kaiser2007). The ever-growing number-crunching capacity of the cortical machinery delivers to the conscious brainstem integrator (“me”) the multichannel feed essential for enhanced decision-making in the Darwinian fight for ascendency. The evolutionary expansion of the cortex was not to implement consciousness, but to provide it with content. The cortex also has direct access to routine bodily functions that can run unattended, in an unsupervised (unconscious) mode.

3. (3) Motivation (appetite for food, water, exchange of gametes, etc.) and emotions (experience of fear, pleasure, and affiliation) developed in the limbic brainstem and forebrain as adaptive elaborations of raw consciousness. In time, the evolving limbic cortex added nuance to the universe of motivation and emotions, in the same way the cortical convexity facilitated conscious decision-making in the brainstem core.

The apparent mismatches between the state-of-Φ and the state-of-consciousness will remain tentative until IIT provides ways to verify that Φ indeed changes as anticipated. Beyond that, although there must be necessary and sufficient conditions for consciousness to emerge, “emergence” per se is not a mechanism. We still need to know how impulses and neurotransmitters generate the sense of “me.” This is not a miracle…consciousness ex nihilo. The feel of missing something also characterizes the irrefutable argument that consciousness is illusory; that working backward from brain state at t = present, given that all prior inputs are already determined, leaves no place for agency, my equally irrefutable feeling that “I exist.”

But there is a third way, the possibility of an intellectually satisfying answer that nobody has come up with yet. Indeed, might the “hard problem” not be hard in the sense of being super-complex, but rather requiring a straightforward concept not yet articulated. After all, history is full of game-changing epiphanies from Darwinian evolution as the explanation of life, to displacement of water for measuring the volume of a golden crown. The physical substrate of consciousness plays out in our brains, it is clearly evolutionarily adaptive (granted that explaining why is a challenge), and, considering the octopus, it may even have evolved independently more than once. Perhaps, the answer will appear during our lifetimes, perhaps in a dream, perhaps tonight.