Perhaps the most powerful contrast conditions for studying a neural correlate of a state of consciousness are comparisons between wakefulness and slow-wave sleep, as well as between wakefulness and general anaesthesia. A notable clinical contrast is a comparison between vegetative state and minimally conscious state patients. Furthermore, new paradigms are available for comparing consciousness with unconsciousness when an overall physiological state of the brain is controlled, such as dreamless vs. dreamful non-rapid eye movement sleep (NREM sleep; Noreika et al. 2009; Siclari et al. 2013) or dreamless vs. dreamful anaesthesia (Noreika et al. 2011). Let us examine several exemplary papers that compare an overall stream of consciousness with its absence.
Sitt et al. (2014) studied auditory-evoked potentials and endogenous fluctuations of EEG signal in 75 vegetative state and 68 minimally-conscious patients. None of the studied evoked potentials (P1, MMN, P3a, P3b, CNV) were able to discriminate patient groups, indicating that task-dependent brain activity does not necessarily distinguish between conscious and unconscious states. Contrary to this, analyses of spontaneous EEG activity showed that unconscious patients had higher power of delta and lower power of theta and alpha oscillations, especially over parietal regions. Furthermore, EEG complexity indices derived from the compressibility of a sequence of data points indicated increased signal complexity over the parietal region in the minimally conscious patients compared to the vegetative state patients. Finally, electrode connectivity measures derived from information theory showed that vegetative-state patients had lower-weighted symbolic mutual information exchange in the range of theta and alpha oscillations than minimally-conscious patients. Interestingly, none of the EEG connectivity measures in the gamma frequency range, including phase lag index and imaginary coherence, could discriminate patient groups, coinciding with other independent observations that gamma synchrony does not necessarily differentiate conscious and unconscious states of the brain (Castro et al. 2013; Steriade et al. 1996).
The finding that the presence of consciousness is associated with an overall complexity of EEG signal and the magnitude of inter-electrode information exchange (Sitt et al. 2014) seems to support the information integration theory of consciousness (Tononi 2012), which predicts that consciousness depends on information complexity and integration in the system. The information integration theory was recently tested by Casali et al. (2013), who investigated the consciousness-related electrodynamics of the distributed cortical networks in a wide range of states of (un)consciousness, including wakefulness (eyes open, eyes closed), sleep (NREM sleep, REM sleep), anaesthesia (midazolam, xenon, propofol), and consciousness disorders (locked-in syndrome, minimally conscious state, patients who have emerged from a minimally conscious state, vegetative state). In a series of experiments, transcranial magnetic stimulation (TMS) pulses were delivered to different cortical sites, which perturbed spontaneous EEG activity (Massimini et al. 2010). Complexity of such TMS-induced EEG perturbations was then calculated, and its index successfully differentiated the states of consciousness and unconsciousness, even at the individual participant’s level (Casali et al. 2013). As predicted, the presence of consciousness was associated with a higher level of information complexity.
In these and similar experiments, the contents of consciousness were not systematically manipulated or controlled for, and conscious participants probably underwent very diverse experiences. Consequently, the reported EEG complexity as the NCC seems to be independent of particular phenomenal contents, and it may reflect some structural aspects of the whole stream of unified subjective experiences. It seems that phenomenal consciousness emerges in a state of the brain that is capable of generating the required level of information complexity and integration. As requested in the previous sections, such an NCC is stable in time and does not depend on an experience isolated from the rest of phenomenal space. Arguably, this type of study tackles the fundamental unity of consciousness much more directly than typical paradigms for studying the selected contents of consciousness. However, approaching one side of the bridge takes us further away from the other side, and the better characterization we have of the neural mechanisms of the state of consciousness, the less we can say about the neural mechanisms of particular contents of consciousness. For instance, the perturbational complexity index can differentiate conscious and unconscious states, but it is extremely insensitive when it comes to distinguishing between different contents of consciousness. For instance, the values of the complexity index did not systematically differ between the “eyes closed” and “eyes open” conditions in the standard waking state (Casali et al. 2013; Noreika 2014). Arguably, any NCC that cannot distinguish between experiences occurring in the “eyes closed” and “eyes open” conditions cannot be fully satisfactory, as the quality of subjective experiences is the core of the scientific problem of consciousness. Yet even though informational complexity does not reflect qualities of phenomenal contents, it is a promising candidate for an NCC of the background properties of consciousness that enable the emergence of subjective experiences and/or necessitate their structural unity.
We are thus left with studies of the contents NCC, such as focusing on the gamma synchrony, and studies of the state NCC, such as focusing on the information complexity. The first group of studies seems to explain the neural binding of concrete selected contents of consciousness, but it does not have a capacity to address the unity of consciousness. The second group seems to capture neural processes involved in the whole stream of consciousness, but it ignores differentiation or phenomenal diversity of consciousness. Ideally, research into the NCC would combine both of these complementary approaches. Unfortunately, a systematic combination of the contents- and states-focused paradigms is almost never tested in cognitive neuroscientific studies of consciousness.
The combined contents-states paradigm would contrast baseline and altered states, or consciousness and unconsciousness, or the transition between the two, while participants carry out experiments that tackle the neural mechanisms of the contents of consciousness. For instance, one could study binocular rivalry while participants lose consciousness in response to an anaesthetic agent. This could, for instance, provide data to investigate how global gamma synchrony as a correlate of the binding and transfer of new contents to awareness depends on or interacts with a changing level of neuronal information complexity. Another promising avenue is research into awareness-related performance in the transition from wakefulness to sleep (Goupil & Bekinschtein 2011). In a recent attempt, Bareham et al. (2014) demonstrated that healthy individuals show neglect-like loss of awareness of the right side of their space in a drowsy state of consciousness. Thus, spatial awareness and unity seem to depend on the state of alertness, as defined by the relative amplitude of theta and alpha oscillations, which confirms that the contents are not wholly independent of the state. That is, despite the external physical stimuli and environment remaining stable, phenomenal contents may appear, disappear, or reorganise depending on the overall state of consciousness. More such studies are expected to be carried out in future, aiming to integrate the content NCC, the state NCC, and altered states of consciousness research programs under one unified framework of the content-state NCC research.