The study of sleep and wakefulness can serve as a pivotal test case to inform debates about the nature of consciousness. We propose that inconsistencies and paradoxes between the position put forth by Merker et al. and Tononi's (Reference Tononi2008) integrated information theory (IIT) of consciousness can be understood in terms of a pervasive false dichotomy between these two states. We argue that neither sleep nor wakefulness is a pure state, and that they fall along a multidimensional continuum, just as consciousness must.

Although many researchers adhere to the traditional conception of human sleep as a global process occurring uniformly across the whole brain, studies of sleep-disordered patients (e.g., Terzaghi et al., Reference Terzaghi, Sartori, Tassi, Didato, Rustioni, LoRusso and Nobili2009) as well as of healthy sleepers (e.g., Nir et al., Reference Nir, Staba, Andrillon, Vyazovskiy, Cirelli, Fried and Tononi2011) indicate that certain populations of neurons and functions can effectively be “asleep” (offline) without us being aware that they are. In fact, there is growing evidence that sleep can be governed by local phenomena (D'Ambrosio et al., Reference D'Ambrosio, Castelnovo, Guglielmi, Nobili, Sarasso and Garbarino2019), that sleep and wakefulness can coexist simultaneously in different brain areas (e.g., Desjardins et al., Reference Desjardins, Carrier, Lina, Fortin, Gosselin, Montplaisir and Zadra2017; Krueger, Nguyen, Dykstra-Aiello, & Taishi, Reference Krueger, Nguyen, Dykstra-Aiello and Taishi2019) and that these interactive processes may give rise to subjective experiences such as attentional lapses and mind wandering (Andrillon et al., Reference Andrillon, Windt, Silk, Drummond, Bellgrove and Tsuchiya2019).

Hence, sleep and wakefulness are not mutually exclusive and the increasingly accepted view of “local sleep” must be distinguished from “global” sleep. Moreover, subjective impressions of sleep and wakefulness are often wrong, such as in the fascinating case of sleep misperception whereby people report being awake during large parts of the night although polysomnographic recordings indicate the presence of unambiguous and persistent sleep (Castelnovo et al., Reference Castelnovo, Ferri, Punjabi, Castronovo, Garbazza, Zucconi and Manconi2019). In addition, one recent laboratory study of good sleepers and insomnia patients with sleep misperception found that the subjective ratings of sleep depth (i.e., being “sound asleep”) reported by both groups was positively correlated with the recall of dream-like conscious experiences prior to being awakened (Stephan, Lecci, Cataldi, & Siclari, Reference Stephan, Lecci, Cataldi and Siclari2021).

In the case of lucid dreaming (being aware that you are asleep and dreaming, while still inside the dream), it is well established that cognitive functions such as working memory, planning, and metacognition can come “online” while the dreamer is still in rapid eye movement (REM sleep (Baird, Mota-Rolim, & Dresler, Reference Baird, Mota-Rolim and Dresler2019), although they may not be functioning in their normal ways, leading to misrememberings, timeline, and novelty detection failures. Lucid dreamers can use volitional eye movements (detected by electrooculography [EOG] recordings in the sleep lab) whereas in REM sleep to communicate the fact that they are aware that they are dreaming to experimenters in the laboratory and, as shown in one recent multicentre study, even engage in two-way communication with experimenters while asleep and dreaming in unequivocal REM sleep (Konkoly et al., Reference Konkoly, Appel, Chabani, Mangiaruga, Gott, Mallett and Paller2021).

The co-occurrence of sleep and wakefulness can also be observed in certain parasomnias, such as sleepwalking, during which patients may engage in complex somnambulistic behaviours such as getting dressed, cooking, playing a musical instrument, climbing a ladder, or driving an automobile, even if showing signs of confusion, poor judgement, and misguided threat perception (Zadra, Desautels, Petit, & Montplaisir, Reference Zadra, Desautels, Petit and Montplaisir2013). In addition, not only are sleepwalkers often aware of their immediate physical environment during an episode, but they can also navigate their surroundings, manipulate objects, interact with other people, and subsequently recall phenomenological contents broadly consistent with the evinced behaviours during their episodes.

These examples, and others like them, support the idea that the simultaneous activation of localized cortical and subcortical networks involved in sleep and wakefulness can occur across a broad range of “conscious” states that transcend the artificial demarcation between sleep and wakefulness.

The notion that consciousness is a unitary thing or process has been challenged by Dennett and Kinsbourne (Reference Dennett and Kinsbourne1992) in this journal, using the case of subjective timing perception. As did those authors, we reject the “Cartesian Theater” model that posits a place, or circuit in the brain where all the information we have “comes together” for some form of mental viewing on a virtual screen in the theatre of the mind, a model that requires an infinite regress of homunculi (if “I” am observing consciousness in my brain, does the “I” that is observing also have a small brain that projects onto its own screen? ad infinitum). If all the tasks of the brain have to be fed into a “consciousness subsystem,” its own activity will have to be decomposed into the activity of further subsystems that begin to duplicate the tasks of the “nonconscious” parts of the whole brain.

Selfridge's (Reference Selfridge1959) influential pandemonium model argued for a plausible account for pattern matching, consistent with Hebbian learning. The basic tenet of pandemonium, that special purpose circuits collect and process information, and the one with the strongest activation “wins” over our consciousness impressions, can be extended to all of perception and cognition, including higher cognitive processes such as decision making. Thus, a neurobiological account of consciousness per se may be nothing more than a snipe hunt, or a search for “where” gravity exists in the room we are now sitting in: it doesn't exist in a single place, it exists as an interactive process among various elements (Luu, Kelley, & Levitin, Reference Luu, Kelley, Levitin and Grossenbacher2001), as shown by resting state functional magnetic resonance imaging (fMRI) (e.g., Atenas and Bustos, Reference Atenas and Bustos2019). Studies of consciousness disordered patients (e.g., Yu et al., Reference Yu, Chen, Zhang, Liu, Tan and Luo2021) reveal classifications based on fMRI activation that can readily distinguish vegetative state from minimally conscious state (MCS) patients and healthy controls, and yet, intriguingly, vast regions of the brain required for perception and cognition did not distinguish these states.

In sum, many of the contradictions between Merker et al.'s account and Tononi's IIT account can be resolved by employing these concepts alongside more precise definitions of “consciousness,” including distinguishing local consciousness from general consciousness and meta-consciousness (awareness of being conscious) – all states that occur not just during wakefulness but also sleep, as during lucid dreaming and certain parasomnias.