Metacognitive Awareness Throughout Sleep and Wake
Metacognitive Awareness
Throughout Sleep and Wake
13-10-2015
Valentina Perdomo
Student number: 10004297 Supervisor: dr. Simon van Gaal Co-assessor: dr. Heleen Slagter
MSc in Brain and Cognitive Sciences, University of Amsterdam Cognitive Neuroscience
Literature Review word count: 11421 Abstract word count: 250
Table of Contents
Abstract 3
Metacognitive Awareness Throughout Sleep and Wake 4
Metacognitive Awareness 6
REM Sleep Lucid Dreaming and Metacognitive Awareness 13
Metacognitive Awareness: A Trait Preserved over the Sleep/Wake Cycle 20
A Relationship between REM lucidity and Wake State Metacognitive Awareness 23
NREM Sleep Metacognitive Awareness: The Missing Link 31
Conclusion & Discussion 38
Abstract
Many benefits are emerging of sleep on wake-time functioning, supporting the importance of sleep. Notwithstanding, sleep is often disregarded; as for many it merely represents a lackluster state of relative unconsciousness. If the phenomenology of sleep could be experienced in a more conscious manner and have this better investigated, this could lead to a gain of interest in this altered state of consciousness. One important facet of consciousness that is believed to be lost during sleep is metacognitive awareness. This literature thesis reviewed whether metacognitive awareness can be maintained throughout the sleep/wake cycle. From the discussed literature, it became evident that sleep need not be characterized by a loss of metacognitive awareness. Lucid dreaming serves as a well-researched model of metacognitively aware REM-sleep. Furthermore, there appears to be a relationship between waking state metacognitive awareness and the ability to lucid dream. Meditators are more likely to be lucid dreamers; lucid dreamers perform better than the normal population on a variety of complex cognitive tasks and also report feeling more in control of their waking state. Furthermore, lucid dreamers have greater gray matter volume in regions implied in metacognitive awareness. This suggests that lucid dreaming reflects a trait of greater metacognitive awareness, rather than the commonly believed viewpoint that lucid dreaming is an intermediate state between wake and sleep. In addition, there is evidence for metacognitive awareness during NREM sleep. These findings support the notion that metacognitive awareness is a trait that can be maintained throughout the sleep and wake cycle.
Metacognitive Awareness Throughout Sleep and Wake
Many benefits are emerging of sleep on cognition and health, supporting the importance of sleep for waking state functioning. From a cognitive perspective, sleep has been implied in improving memory consolidation, supporting future memory encoding and generating sleep-related insights over waking state problems (Talamini, Nieuwenhuis, Takashima, & Jensen, 2008; Van Der Werf et al., 2009; Wagner, Gais, Haider, Verleger, & Born, 2004). From a health perspective, sleep has been linked to improved immune system, metabolism and endocrine functioning and better cardiovascular health (Dinges, Douglas, Hamarman, Zaugg, & Kapoor, 1995; Knutson, 2010; Spiegel, Leproult, & Van Cauter, 1999). The discernible importance of sleep, however, is not reflected in the sleep hygiene of our society. In the U.S. population, an average night’s sleep has decreased by one to two hours over the course of the last four decades (Kripke et al., 2002, cited in in Calamaro, Mason, & Ratcliffe, 2009). For many professionals, students and parents alike, sleep is seen as nothing more than a sad necessity, a waste of valuable hours that could otherwise be better spent on conscious waking pursuits. Indeed, “I’ll sleep when I’m dead” has become somewhat of a mantra for many busy individuals as a defense against criticisms of a sleep-deprived lifestyle. This is not far-fetched, as in contrast to primitive cultures; sleep to the modern West represents a lackluster state of relative unconsciousness (Lincoln, 2003; Massimini, 2005; Steriade, 2003). Sleep mentation (mental activity during sleep) still carries the connotation of being deficient to waking state mentation and the interest in this altered state of consciousness appears to be all but lost (Rechtschaffen, 1978). In a by now seminal article, Alan Rechtsschaffen (1978) described dreams as a state of single-mindedness and isolation:
Single-mindedness of dreams would be to say that dream consciousness, at least on a manifest level, is isolated from other systems of consciousness, i.e., reflection, voluntary control, other images, etc. This isolation may be just one manifestation of a more generalized isolation of dream consciousness, not only from other systems of consciousness, but from stimulus input, autonomic activity, organismic state, and motor output as well (p. 103).
However, the notion of sleep as an unaware and unconscious affair might be more due to cultural influences that are already apparent in waking state cognition. If the phenomenology of sleep could be experienced in a more aware manner and have it be better researched, this could feasibly lead to regaining of interest in this altered state of consciousness and thus also lead to people taking the importance of sleep more seriously.
This literature thesis reviews whether metacognitive awareness (awareness over one’s own mental processes and state) can be maintained throughout the sleep/wake cycle. More specifically, the commonly accepted notion that sleep is a relatively unconscious state characterized by a lack of metacognitive awareness as compared to the waking state is disputed. First, the construct of metacognitive awareness will be elaborated upon and the consensus of sleep as a metacognitively unaware process will be reviewed, focusing on the neural substrates that are thought to underlie this loss of metacognitive awareness. Second, this view of sleep being a metacognitively unaware process will be brought into question by means of previous literature on REM sleep lucid dreaming, which indicate that sleep need not be characterized by a loss of awareness that one is residing in a sleeping state. Third, individual differences will be considered as to how they may play a role in maintaining metacognitive-awareness throughout sleep. Individual differences already
present during the wake state would support the view that lucid dreaming is not an anomalous state of sleep posited as an intermediate state between wake and sleep, but simply part of a differential form of cognitive functioning that can be preserved across the sleep wake cycle, at least for REM sleep and wake. Lastly, literature on NREM metacognitive awareness will be reviewed to support the notion that metacognitive awareness can be maintained through all stages of the sleep/wake cycle.
Metacognitive Awareness
Metacognitive awareness in its strictest sense refers to having awareness over first-order experience or cognition (Jonathan W. Schooler, 2002). It is part of higher-first-order consciousness and allows access to the content of primary consciousness (Edelman, 2003; Seth, Baars, & Edelman, 2005). Primary consciousness refers to the direct experience of sensory and motor events. Primary consciousness is believed to be present in majority of animals. In contrast, higher-order consciousness involves the referral of the contents of primary consciousness to interpretative semantics, including a sense of self. Therefore, the presence of metacognitive awareness may be
constitutively required for advanced forms of self-reflective awareness. It is assumed to only be present in certain hominoids and presents itself in the richest form in humans (Edelman, 2003).
Metacognitive awareness is likely to be subserved by the anterior frontal cortex, or more specifically, the FrontoPolar (FPC) and DorsoLateral Prefrontal Cortex (DLPFC) (Christoff & Gabrieli, 2000). The FPC is the most anterior part of the prefrontal cortex and corresponds to Brodmann Area (BA) 10 (Brodmann, 1909). It underwent a significant expansion and reorganization during hominid evolution; the human FPC doubled in relative size to those of Bonobos and Chimpanzees, in
reorganization (Katerina Semendeferi & Damasio, 2000; K. Semendeferi et al., 2010). The DLPFC is a functionally, rather than an anatomically defined brain structure (Cieslik et al., 2012). It consists of Brodmann area 46 and 9 and is the most recently evolved part of the human brain (Nelson, Collins, & Luciana, 2001). It is also known to undergo an extremely prolonged period of maturation that lasts until early
adulthood (Brodmann, 1909; Hill et al., 2010) (see Figure 1 for a visual representation of these regions).
Figure 1. A frontal and lateral view of the brain regions of interest. The FrontoPolar Cortex (FPC) is
Brodmann Area (BA) 46 and 9. Figures compiled and adapted from BodyParts3D/Anatomography (n.d.). Licensed under the Creative Commons Attribution-Share Alike 2.1 Japan.
These two cortical regions are hypothesized to form a system specialized for the monitoring of information held in working memory, which can provide the subjective experience of metacognitive awareness over this information. It is proposed that there may be a hierarchical distinction in a rostrocaudal direction between the FPC and the DLPFC. The DLPFC may be sufficient for the monitoring of recently or currently presented externally generated information, whereas the FPC is additionally recruited when monitoring of internally generated information not present in the environment is needed (Christoff & Gabrieli, 2000). For a more extensive review of the distinction between these two types of metacognitive awareness, see Christoff & Gabrielli (2000).
In addition, a central premise in recent theorizing about metacognitive awareness during the waking state is that it corresponds to a discontinuous process whereby individuals periodically notice the current content of their mind (Jonathan W. Schooler et al., 2011). This perspective is informed by mind-wandering research. Mind-wandering, also known as stimulus-independent thoughts, is characterized by a lack of metacognitive awareness (Fox, Nijeboer, Solomonova, Domhoff, & Christoff, 2013). Mind-wandering research primarily entails the combination of self-catching measures of the mind-wandering state with experience sampling probes during a task (e.g. reading) (Jonathan W. Schooler et al., 2011). The self-catching measure asks participants to press a response key every time they notice for themselves that they have been mind wandering and thus redirect their attention to the task at hand. This provides a straightforward assessment of the number of mind-wandering episodes that reached metacognitive-awareness. With experience sampling probes, subjects are asked if they were mind-wandering in the period directly before the probe. By
contrast to the self-catching measure, the experience sampling probes allow an assessment of the amount of mind wandering that is actually taking place. When such probes catch people mind-wandering before they notice it themselves and redirect their attention to the task, the experimenter is able to quantify the relative amount of mind wandering that the individual is unaware of. On average, 15% of the time during experience sampling probes, people were caught mind-wandering, which they had not gained metacognitive awareness over (Jonathan W Schooler, 2004). However, given that the subjects were within an experimental setting and knew that their level of awareness over their mind-wandering was being measured, in addition to being given the explicit instructions to observe their mind-wandering, it may have led to these subjects being more vigilant over their mental state than during daily life.
In a related paradigm, where subjects were probed during their day-to-day activities to indicate whether or not they were mind-wandering, mind wandering occurred in close to 50% of the samples taken (Killingsworth & Gilbert, 2010). The amount of mind-wandering also varied per task, however, it occurred in at least 30% of the samples taken during every activity except making love (Killingsworth & Gilbert, 2010; McVay, Kane, & Kwapil, 2009). Given that these studies were based in a naturalistic setting and did not use a self-catching measure to probe the extent that subjects became aware of their mind-wandering, it is likely that these percentages are closer day-to-day levels of mind-wandering. This has as implication that a lack of metacognitive awareness may be present in close to 50% of the waking state in the normal population.
There are, however, also trait-level differences present when it comes to noticing the content of one’s mind. For example, mindfulness and meditation training cultivates moment-to-moment awareness of the self and environment (Wallace &
Goleman, 2006). To this extent, mindfulness training heightens metacognitive awareness and decreases mind-wandering (Austin, 1998; Ortner, Kilner, & Zelazo, 2007). Meditation practice is also greatly governed by the skill to observe one’s own ongoing mental processes and entails the continuous monitoring of mental activity (metacognitive awareness) (Lutz, Slagter, Dunne, & Davidson, 2008).
On a neural level, greater DLPFC and FPC activation are commonly found at the onset of a meditation session as well as during meditation (Baerentsen, Hartvig, Stødkilde-Jørgensen, & Mammen, 2001; Cahn & Polich, 2006; Farb et al., 2007; Manna et al., 2010; Newberg & Iversen, 2003; Ritskes, Ritskes-Hoitinga, Stødkilde-Jørgensen, Bærentsen, & Hartmann, 2004; Ritskes-Hoitinga, 2004). Greater Gray Matter Volume (GMV) has also been reported in the FPC in meditators of disparate (Vipassana, Tibetan and Brain Wave Vibration) meditative backgrounds, suggesting that even different styles of meditation practices have this region in common (Kang et al., 2012; Lazar et al., 2005; Vestergaard-Poulsen et al., 2009). Structural brain differences for the DLPFC are less commonly reported. This may be due to the DLPFC consisting out of 2 anatomically defined brain regions, namely BA 46 and 9 (see Figure 1). In a study by Lazar et al. (2005), greater GMV was present in BA 9 but not 46 in advanced meditators as compared to non-meditators. A more recent Region of Interest (ROI) study by Hölzel et al. (2008) revealed no structural differences in meditators as compared to non-meditators for the DLPFC (BA 9 & 46). Given that this was a hypothesis-driven study on a brain region comprised of both Brodmann areas, it was not investigated whether each Brodmann area may have shown differences in GMV. It may be that only BA 9 plays a role in metacognitive awareness commonly found in meditators and not BA 46. An alternative explanation may be related to the hierarchical distinction in the rostrocaudal direction between the
FPC and the DLPFC (Christoff & Gabrielli, 2010). Possibly the reason for why structural differences are most evident for the FPC is that monitoring of the internal state, as opposed to monitoring of recently presented information, is more uniquely trained with meditation. This discrepancy aside, these findings suggest that there are also state and trait-level differences present in metacognitive awareness to the extent that these regions are more active during metacognitively intensive tasks such as meditation and that meditative expertise is related to structural increases in GMV.
The above reviewed findings, however, concern waking state metacognitive awareness. Given that on average a third of our lives is spent asleep, it is of importance to understand what occurs with metacognitive awareness during the sleeping state (Biddle & Hamermesh, 1989). Sleep itself can be divided into NREM (N1, N2, N3) and REM sleep following the guidelines of the American Academy of Sleep Medicine (AASM) standards (Medicine & Iber, 2007). When falling asleep, there is a transition from wake to NREM. NREM sleep and REM sleep then alternate through the night in a cyclical fashion. Most slow-wave NREM sleep occurs in the first part of the night with REM sleep episodes becoming longer throughout the night (Medicine & Iber, 2007).
Currently, it is believed that there is a complete fading out of metacognitive-awareness in humans every night as a result of sleep, which is regained upon awakening (Hobson, 2010; Muzur, Pace-Schott, & Hobson, 2002; Steriade, 2003). According to Muzur et al. (2002), the loss of metacognitive-awareness occasioned by the transition from waking to sleep constitutes one of the most universal and robust changes in consciousness that humans can observe in themselves and others. The quality of subjective experience during the sleep stages may differ; mentation reports obtained after REM sleep awakenings tend to be longer, more bizarre, visual, motoric,
and emotionally charged than after NREM sleep, which are reported to be more logical, thought-like, ruminative and linked to current concerns (Fosse, Stickgold, & Hobson, 2004; Hobson & Stickgold, 1994; Kales et al., 1967). However, both states are believed to be characterized by a complete lack of awareness of residing in a sleeping state.
Indeed, the prefrontal cortex shows the greatest change from waking to sleep (Braun, 1997). The transition from waking to NREM is characterized by frontal deactivation as reported in positron emission tomography (PET) studies and quantitative EEG studies (Kajimura et al., 1999; Maquet et al., 1997; Werth, Achermann, & Borbély, 1997). Deactivation increases with the deepening of NREM sleep (Hofle et al., 1997). PET studies have demonstrated that the decline of regional cerebral blood flow during slow-wave activity (SWA) is most prominent in frontal cortical areas (Braun, 1997; Hofle et al., 1997; Kajimura et al., 1999; Maquet et al., 1997; Nofzinger, Mintun, Wiseman, Kupfer, & Moore, 1997). Even though certain frontal regions regain activity (most posterior and medial prefrontal regions) with the onset of REM sleep, the anterior and lateral portions, of which the FPC and DLPFC are comprised of, remain relatively deactivated (Braun, 1997; Maquet et al., 1996).
The hypothesized reason for this, according to Muzur et al. (2002), is that executive cognitive functions, provided by the prefrontal cortex, in particular the DLPFC, are particularly sensitive to the fatigue induced by prolonged waking and that it can “enjoy a passive respite” during sleep (Fuster, 2001; Goldberg, 2002). In addition, it is suggested that it is only by virtue of this respite provided by sleep that the prefrontal cortex recovers its crucial functional competence for use on waking.
To summarize, metacognitive awareness, which is subserved by the FPC and DLPFC, is assumed to be present in its richest form in the human species and
according to mind-wandering research occurs intermittently throughout the day (Edelman, 2003; Schooler et al., 2011). In addition, there are also trait-level differences when it comes to metacognitive awareness and meditation and mindfulness practices are thought to increase it (Austin, 1998; Ortner, Kilner, & Zelazo, 2007). Currently, it is also believed that there is a complete fading out of metacognitive-awareness in humans every night as a result of sleep, which is regained upon awakening (Muzur et al., 2002; Steriade, 2003). This is assumed to be part of the restorative component of sleep.
However, even though the current consensus may be that sleep is characterized by a loss of metacognitive awareness and prefrontal cortical activity, it may still be possible that this is strictly one method of sleeping, especially in the light that there are also trait-level differences in the cognitive skill of metacognitive awareness for the waking state (Austin, 1998; Lutz et al., 2008; Ortner et al., 2007). Therefore the statement by Muzur et al. (2002) that a loss of metacognitive-awareness during sleep is a universal and robust phenomenon could potentially be nuanced.
REM Sleep Lucid Dreaming and Metacognitive Awareness
Even though Muzur et al. (2002) posit sleep as a phenomenon that is universally characterized by a fading out of metacognitive awareness, there are exceptions that could be made. Lucid dreaming provides such an exception of sleep that is not characterized by a robust loss of metacognitive awareness and prefrontal activation. A high level of metacognitive awareness while in REM sleep characterizes lucid dreams, i.e. dreaming while being conscious that one is dreaming (Laberge, Nagel, Dement, & Zarcone, 1981; Payne, 2014). Lucid dreams are thought to arise exclusively from non-lucid dreams in REM sleep in which the subject questions
whether he or she is dreaming as a result of for example noticing bizarre elements in the dream.
In the modern West the first notable mention of lucid dreaming is credited to Frederik van Eeden, a Dutch Psychiatrist who coined the term (Van Eeden, 1913). Van Eeden (1913) describes the phenomenon of lucid dreaming as the following:
In these lucid dreams the reintegration of the psychic functions is so complete that the sleeper remembers day-life and his own condition, reaches a state of perfect awareness, and is able to direct his attention, and to attempt different acts of free volition. Yet the sleep, as I am able confidently to state, is undisturbed, deep and refreshing (p. 5).
However, it was not until the early 1980’s that the first substantial proof for this phenomenon was found when Stephen LaBerge trained subjects to perform a certain pre-agreed upon amount of voluntary horizontal eye saccades (left-right eye movements) to signal to the experimenter of an ongoing lucid dream (LaBerge et al., 1981).
Recently, with modern neuroimaging techniques, it was also found that greater levels of metacognitive-awareness during sleep is indeed related to an increase of prefrontal activity in the same regions implied in waking state metacognitive awareness (Dresler et al., 2012; Voss, Holzmann, Tuin, & Hobson, 2009). Voss et al. (2009) trained a group of 20 undergraduate Psychology students in lucid dreaming. After four months of weekly training sessions, six subjects claimed to experience lucid dreams regularly (three lucid dreams a week) and were invited for laboratory EEG measurements. Three of the six subjects managed to obtain lucidity within a laboratory setting. Current Source Density (CSD) transformation was performed on EEG data for better source localization. An increase of 40 Hz. Gamma power was
found during lucid REM sleep for the frontal and frontolateral regions of the brain as compared to non-lucid REM sleep (see Figure 2). Previously, a close spatial correspondence has been found between regions of fMRI activations and recording sites showing increases in EEG gamma power (Lachaux et al., 2007). Therefore, greater gamma power of the frontal and frontolateral regions likely represents greater activation of these frontal and frontolateral brain regions (FPC and DLPFC).
The relationship between lucid dreaming and greater prefrontal activation became more evident in an fMRI study by Dresler et al. (2012). In this study, one of the four subjects measured managed to lucid dream in the fMRI scanner during two subsequent whole night sleep-recordings. As per definition, dream reports of these two lucid dreams indicate the lucid dreams started out of rather confused sleep mentation (non-lucid dreaming) without clearly memorized content. fMRI analyses were within-subject based and were done on these 2 episodes of lucid REM sleep as compared to non-lucid REM sleep. Greater activation was found for lucid REM sleep as compared to non-lucid REM sleep in various cortical regions, of which the FPC and DLPFC. This provides direct supporting evidence of the role of these brain regions in metacognitive awareness during sleep.
Figure 2. Lucid (LUCID) dreaming was found to have higher 40 Hz. power than non-lucid (NLUCID)
dreaming but lower 40 Hz. power than waking state Eyes-Closed-Rest (ECR). This figure illustrates 40 Hz. power in lucid dreamers for the waking state, lucid dreaming and non-lucid dreaming as measured with 19-channel EEG after CSD transformation. Adapted from “Lucid dreaming : a state of
consciousness with features of both waking and non-lucid dreaming” by Voss et al., 2009, Sleep, 32(9), 1191. Copyright 2015 by Associated Professional Sleep Societies, LLC.
A follow up study by Stumbrys, Erlacher, & Schredl (2013) aimed to manipulate the activation of the DLPFC during REM sleep through transcranial Direct Current Stimulation (tDCS) to increase dream lucidity. Nineteen participants randomly received either tDCS for 10 minutes or sham stimulation during each REM period starting with the second REM period. According to the participants’ self-ratings, tDCS over the DLPFC during REM sleep increased lucidity in dreams. Their effects, however, were not very strong and were found only in frequent lucid dreamers. A reason for this may be due to their participant selection criteria. Several times during the laboratory habituation night, tDCS was shortly applied during REM sleep. If after each application the participant was awakened, he or she was considered to be too sensitive to tDCS and was withdrawn from further participation in the study. This may have led to a selection bias given the intrusive nature of tDCS. For the Voss et al. (2009) study, it was apparent that all six lucid dreamers were very sensitive to light and sounds and were easily awakened as a result of lucid dreaming induction methods utilizing visual and auditory stimuli. It may be that lucid dreamers are particularly sensitive to external disturbances during sleep, which might extend to tDCS stimulation. This may have inadvertently led to the removal of subjects who may have been more prone to lucid dreaming.
Voss et al. (2014) obtained better results utilizing a similar design. The greatest difference between these two studies was the use of Transcranial Alternating Current Stimulation (tACS), as opposed to tDCS. With tACS, rather than applying a direct current to brain activity, sinusoidal current can be applied. This offers the benefit of manipulating specific EEG frequency-bands to test how they affect brain activity and metacognitive awareness during sleep. In addition, the current strength
was chosen well below sensory and below phosphene threshold, and smooth ramp-up/ramp-down phases were used to avoid awakening of the subject. Therefore, subject removal was not required. It was hypothesized that there would be a causal relationship between 40 Hz. gamma power on lucid dreaming, given that Voss et al. (2009) found greater 40 Hz. gamma power during lucid dreams. Twenty-seven non-lucid dreamers were measured up to four non-consecutive nights. Subjects were allowed to sleep undisturbed until 3:00 a.m., as it is known that early morning sleep, unlike late night sleep, is more conducive to lucid dreaming due to REM sleep being more prominent. Starting at 3 a.m., and following at least 2 minutes and maximally 3 minutes of uninterrupted, arousal-free REM sleep, tACS was applied fronto-temporally (over the DLPFC) at various frequencies (2, 6, 12, 25, 40, 70 and 100 Hz.) and a sham condition (simulated stimulation, but no current flow) in a double blind design (see Figure 3 for an illustration of the research design). After which subjects were awakened and asked to rate dream consciousness based on their subjective experience and a validated lucidity scale. Subjective experiences of lucid dreams were most prominent during 40 Hz. Gamma power stimulation, which were reported by subjects to have occurred 77% of the time. Interestingly, the extent to which tACS stimulation induced increases in frontal gamma power correlated positively with the degree to which participants experienced lucid dreaming. This provides a direct causal relationship between frontal Gamma power, which is known to be linked to greater activation of the frontal regions, and metacognitive awareness during REM sleep (Lachaux et al., 2007; Voss et al., 2014).
F igure 3. Double-blind repeated measures tACS stimulation (30 s duration each) in REM sleep over 4
non-consecutive nights. Frequency of stimulation (sham, 2, 6, 12, 25, 40, 70, and 100 Hz) was counterbalanced across subjects and across nights. Reprinted from “Induction of self awareness in dreams through frontal low current stimulation of gamma activity” by Voss et al., 2009, Nature
neuroscience, 17(6), 810-812. Copyright 2003 by the Nature Publishing Group.
In conclusion, it becomes apparent that the prefrontal de-activation with its loss of metacognitive awareness need not be the only way of sleeping. Lucid dreaming provides a model of metacognitively aware sleep that is characterized by greater activity the FPC and DLPFC and greater Gamma power of the frontal and frontolateral regions as compared to non-lucid dreaming (Dresler et al., 2012; Voss et al., 2009). For the latter, this is also indicative of greater activation of the underlying cortical regions (Lachaux et al., 2007). Furthermore, a direct causal relationship can be inferred with the studies by Stumbrys et al. (2013) and Voss et al. (2014), which found that stimulation of the frontal regions (DLPFC) during REM sleep led to increases in metacognitive awareness.
However, if the definition of sleep as a metacognitively unaware phenomenon is upheld, lucid dreaming becomes a bit of a paradox. Likely for this reason Voss et al. (2009; 2014), state that lucid dreaming represents a dissociated state containing both wake and dreaming features. In this case, the relevant waking feature would be the higher frequency power in the frontal regions accompanied with a greater level of metacognitive awareness while the dreaming features relate to the neural markers of
REM sleep (increases in delta and theta power). In a way, there is no other way than to define lucid dreaming as such if sleep is defined as a state deficient of metacognitive awareness compared to the waking state. Most likely due the small sample size (N=3), Voss et al. (2009) did not compare these subjects to a control population on the measurements obtained. This had as consequence that potential differences between these two populations were not directly investigated. Even so, of the 20 subjects trained in lucid dreaming, only six of those that were part of this study became regular lucid dreamers. This implies a certain level of predisposition for lucid dreaming. It may be that these lucid dreamers already have greater high-frequency power in the frontal and frontolateral regions during the waking state and are therefore more likely to dream lucidly. In such a case, this greater high-frequency power would be a trait maintained throughout wake and sleep and would not represent a hybrid state, but rather a phenomenon that is relatively independent of the sleep/wake cycle.
Potentially, the difference between non-lucid REM and lucid REM within a subject is strictly related to this intermittent process of metacognitive awareness already present during the waking state (Jonathan W Schooler, 2004; Jonathan W. Schooler et al., 2011). Metacognitive awareness in the normal population is indeed believed not to be a continuous process, but a discontinuous one where people periodically notice the contents of their minds. Thus it may be that during REM sleep, individuals may occasionally notice that they are dreaming and have this lucidity fade out again. Consequently, if this were true, lucid dreaming would lose its appeal as this anomalous state between wake and sleep and would actually become part of the regular attentional dynamics of metacognitive awareness as is apparent in the normal population. Hypothetically, due to individual differences in metacognitive awareness, some may not become aware that they are residing in the dreaming state at all, while
others may potentially be able to maintain this awareness throughout the night (and day).
Furthermore, the decrease of gamma power in lucid dreaming as compared to the waking state is taken as proof for lucid dreaming being an intermediate state between wake and sleep. However, this may also have been related to the detail that the three subjects measured were novice lucid dreamers (four months of training). It may be that with people more familiar with lucid dreaming may show wake-state levels of gamma power during REM sleep. It is therefore yet unclear if lucid dreaming is indeed a dissociated state containing features of both wake and dreaming (intermediate state between wake and sleep) or whether it simply entails greater trait-level metacognitive awareness in lucid dreamers that is relatively preserved across the sleep and wake cycle due to individual differences (see Figure 4).
F igure 4. Lucid Dreaming as a dissociated state (left) or as differential functioning that is preserved
across the sleep-wake cycle (right). The left figure illustrates the idea by Voss et al. (2009) that lucid dreaming is a dissociated state containing features of both wake and dreaming (intermediate state between wake and sleep), i.e. a higher level of “awakeness” is maintained when falling asleep. In this case “awakeness” represents the level of metacognitive awareness maintained when falling asleep. The right figure illustrates an alternative viewpoint, which states that lucid dreaming may actually entail differential functioning preserved across the sleep and wake cycle. In this case, it may that lucid dreamers were strictly more “awake” to begin with (differential baselines of metacognitive awareness for the wake state).*
* The slope of the right figure indicates that there is a general decrease in metacognitive awareness as a result of dreaming. This is strictly for illustrative purposes and ease of understanding. Kahan & LaBerge (1994) maintain that sleep mentation is not deficient to waking state mentation and it may be that for the figure on the right, both slopes would be best represented with a denominator of 0.
Metacognitive Awareness: A Trait Preserved over the Sleep/Wake Cycle
In a book on dreaming, Gordon Globus (1987) strikes up an intriguing possibility, namely that upon reflection of his life, he bears upon an unreflective experience as lived, and it is just this unreflective experience of a world that is at times indistinguishable across waking and dreaming. To state this in more cognitive terms, Globus remarks upon his lack of metacognitive awareness during the waking state, which in retrospect seems indistinguishable at times to him from his dreams. It may thus be that similar cognitive processes, or better said a lack thereof, are maintained throughout wake and sleep. To illustrate this notion, parallels have been drawn between mind-wandering, which the normal population is purported to reside in close to 50% of their time spent awake, and dreaming (Fox et al., 2013; Killingsworth & Gilbert, 2010). As described above, both states are characterized by a lack of metacognitive awareness, in addition to also containing similar content. In comparing the first-person descriptions to meta-analytic data from numerous functional neuroimaging (PET, fMRI) studies of the default mode network (with high chances of mind-wandering) and rapid eye movement (REM) sleep (with high chances of dreaming), large overlaps in activation patterns of cortical regions were found. In addition, both states are characterized by deactivation of numerous PFC executive regions, with REM sleep exhibiting greater deactivation. The authors suggest that non-lucid dreaming may simply represent an amplified trait (mind-wandering) already present in the waking state. Thus, it may be that populations who have greater metacognitive awareness and experience less mind-wandering may also experience sleep that contains a greater level of metacognitive awareness.
Lucid dreaming serves as a valuable model of sleep that is characterized by a greater level of metacognitive awareness. Likely therefore, the majority of research on
lucid dreamers is based on sleep studies. However, if these subjects are more metacognitively aware during the sleeping state, it is also of interest to investigate if they are also more metacognitively aware of their wake state and vice versa. General trait-level differences in metacognitive awareness present during the waking state in lucid dreamers would provide support for the hypothesis that lucid dreamers do not exhibit a dissociated state of sleep during lucid dreaming, but strictly have greater trait-level metacognitive awareness, which is preserved across the sleep/wake cycle. Furthermore, given the general axiom in lucid dreaming research that lucidity can only be gained during REM sleep dreams, NREM lucidity should also be considered if metacognitive awareness is indeed a phenomenon maintained to a certain extent across sleep and wake states.
After all, the continuity hypothesis postulates that there is considerable continuity across sleep and wake mentation (Domhoff, 1996). This is in sharp contrast to the still popular view, originally proposed by Freud, that the dreaming mind differs greatly from the waking mind (Freud, 1900; Muzur et al., 2002). Proponents of the continuity theory, however, focused primarily on the continuity of content across sleep and wake; their central claim is that dreams reflect an individual’s waking life experiences, concerns, and personality (Domhoff, 1996; Schredl & Hofmann, 2003)/ However, it may be that specific wake-state processes and their individual differences are also better preserved during sleep than currently believed (Fox et al., 2013; Muzur et al., 2002).
Indeed, there is some supporting evidence for this from Laberge’s later work on comparing accounts of dreaming and waking state mentation (Kahan & LaBerge, 1996). Even though not meeting the requirements for lucid dreaming, subjects maintained a much more comparable level of metacognitive awareness to waking
state mentation during dreams than what is commonly believed. With the exception of the dimension of choice (the ability to make conscious decisions), no differences were found in dream reports and comparative wake reports for the occurrence of metacognitive processes (e.g. self-reflective awareness) (Kahan, & Laberge, 1996). This provides supporting evidence of considerable continuity of metacognitive processes across waking and dreaming.
Given that upon closer inspection, the notion of sleep mentation being deficient compared to wake state mentation is not as concrete as is often asserted, it may be that processes like metacognitive awareness are much better preserved across the sleep/wake cycle than previously believed. Therefore, it is of interest to better investigate how REM sleep metacognitive awareness relates to wake and NREM sleep metacognitive awareness.
A Relationship between REM lucidity and Wake State Metacognitive Awareness
Estimates vary, but it is purported that on average ~ 20% of the population are frequent lucid dreamers (i.e. having more than one lucid dream a month) (Erlacher, Schredl, Watanabe, Yamana, & Gantzert, 2008; Erlacher, Stumbrys, & Schredl, 2011). Aside from predisposing factors in certain individuals to lucid dream, lucid dreaming can also be trained. According to Purcell, Mullington, Moffitt, & Hoffmann (1986), lucid dreaming is a cognitive skill that can be increased by attentional techniques learned when awake. As Tholey (1983) states, “one method to achieve lucidity is by the practice of developing a critical self-reflective frame of mind concerning one’s state of consciousness” (p.79). This is also referred to as lucid awareness training (Stumbrys, Erlacher, Schädlich, & Schredl, 2012). Developing a critical self-reflective frame of mind concerning one’s state of consciousness would
require metacognitive awareness into the current state of mind, implying indirect evidence for a relationship between sleep and wake state metacognitive awareness.
If there is indeed a relationship between sleep and wake metacognitive awareness, this relationship should also be reflected in behavioral performance differences in lucid dreamers as compared to non-lucid dreamers. Even though not as thoroughly investigated as REM sleep studies in lucid dreamers, there have been a few studies conducted on the capacity to lucid dream on waking state performance. Lucid dreamers are known to perform better on a variety of complex cognitive tasks, such as the Stroop task, The Remote Association Task, and the Idaho Gambling task (Blagrove, Bell, & Wilkinson, 2010; Bourke & Shaw, 2014; Neider, Pace-Schott, Forselius, Pittman, & Morgan, 2011).
Interestingly, the DLPFC, one of the regions implied in metacognitive awareness (Christoff & Gabrielli, 2000), has been suggested to play a crucial role in performance in all of the tasks above (Brevers, Bechara, Cleeremans, & Noël, 2013; MacDonald, Cohen, Stenger, & Carter, 2000; Qiu et al., 2010). In a by now keystone article, Mcdonald et al. (2000) found that individuals who showed the most DLPFC activation after the color-naming instruction on the Stroop task showed the smallest Stroop interference effect. In addition, solving compound remote associate problems and performance on the Idaho Gambling task have also been explicitly linked to the activation of the DLPFC (Qiu et al., 2010; Brevers et al., 2013). Even though differential explanations have been provided for how the DLPFC may play a role in improving performance on each of the tasks listed above, it may be that all of these regions have an overarching function in common of for example greater awareness over incoming stimuli, in addition to greater control over the most adaptive response due to having greater metacognitive awareness (Christoff & Gabrieli, 2000).
In addition, personality traits of lucid dreamers have also been investigated. Lucid dreamers report having a greater internal locus of control, i.e., lucid dreamers feel more in control of their waking life (Blagrove & Tucker, 1994; Blagrove & Hartnell, 2000). Given that lucid dreaming is also characterized by a greater ability to exert control over dreams, this offers additional support for the continuity hypothesis in which lucid dreamers report feeling more in control of their sleep and wake states (Domhoff, 1996). Similarly, Gruber, Steffen and Vonderhaar (1995), found that the personality dimension that best distinguishes frequent from infrequent and non-lucid dreamers on the 16PF personality test is that of subduedness/independence. Lucid dreamers were more initiative taking while infrequent and non-lucid dreamers were more passive and in need of external support. In addition, lucid dreamers were also found to be more socially bold, dominant, experimenting, enthusiastic and warm on the 16PF personality test than non-lucid dreamers. Conversely the opposing poles of these dimensions, which better characterized non-lucid dreamers, describe individuals who are shy, submissive, conservative, restrained, and cool. The authors suggest that these findings strongly suggest that frequent lucid dreamers, characterized by the unusual degree of self-reflection and volitional control they exhibit in the dream state, are also more proficient at the management or control of various aspects of cognitive, emotional, and social functioning while awake.
Potentially, this proficiency at the management of waking state cognition (e.g. performing better than the normal population on a variety of cognitive tasks) and emotion may be relying on the same core cognitive process that gives lucid dreamers greater awareness and control in their REM sleeping state. Indeed, metacognitive awareness is known to give people more control over for example their emotions by providing greater awareness into the current emotional state and thus decreasing
emotional reactivity (Farb et al., 2007). In this study, fMRI was employed to compare neural reactivity to sadness provocation in 20 participants completing 8 weeks of mindfulness training and 16 waitlisted controls. Sadness resulted in widespread recruitment of regions associated with self-referential processes along the cortical midline. Participants who had completed the mindfulness training, however, demonstrated a distinct neural response with greater lateralized brain region recruitment consisting of the DLPFC.
Increased metacognitive awareness was also associated in previous studies with reduced endorsement of dysfunctional cognitions following sadness challenge, reduced cognitive processing of negative material in present moment awareness, and an increased willingness to tolerate negative affect (Arch & Craske, 2006; Fresco, Segal, Buis, & Kennedy, 2007; Frewen, Evans, Maraj, Dozois, & Partridge, 2007). It is thus comprehensible how the development of greater metacognitive awareness may incur greater management or control of various facets of wake life (e.g. the regulation of negative affect), which in turn might also allows subjects to more likely achieve states of self-reflection and volitional control while dreaming (Gruber et al., 1995).
If there is consistent greater activation in the brains regions implied in metacognitive awareness for lucid dreamers for both the waking (DLPFC) and sleeping state (DLPFC & FPC), structural neuroplasticity would dictate that this should in its turn also lead to training induced structural changes in the adult human brain (Blagrove et al., 2010; Bourke & Shaw, 2014; Draganski & May, 2008; Dresler et al., 2012; Neider et al., 2011; Voss et al., 2009). In other words, differential performance should also be reflected in differences in GMV, as was apparent in meditators (Lazar et al., 2005; Vestergaard-Poulsen et al., 2009; Kang et al., 2013).
This was investigated in a study by Filevich et al. (2015), in which they sought to explore structural brain differences in lucid dreamers as compared to non-lucid dreamers. Sixty-three participants were median split into a high trait-level and low trait-level lucidity group based on a composite score of self-reported lucid dream frequency and results from the LuCid questionnaire (Voss, Schermelleh-Engel, Windt, Frenzel, & Hobson, 2013 cited in Filevich et al., 2015). Filevich et al. (2015) found greater GMV in the DLPFC and FPC in high trait-level lucid dreamers (BA 9 and 10) as compared to low trait-level lucid dreamers. In addition, these same regions (BA and 10) identified through structural analyses also showed increases in BOLD signal during thought monitoring in both groups, in addition to greater BOLD signal for the high trait-level lucidity group. These two findings support the notion of structural differences in regions implicated in metacognitive awareness being present in lucid dreamers as compared to non-lucid dreamers. The authors suggest that these are the same regions also known to play a role in visual metacognitive awareness and conclude that this raises the intriguing possibility that lucid dreaming might rely on some core metacognitive mechanisms shared across different metacognitive tasks (De Martino et al., 2013, cited in Filevich et al., 2015). However, given the evidence presented in this review, particularly in light of the continuity hypothesis, the notion of some core metacognitive mechanism that is shared across tasks might appear better founded than strictly being presented as an intriguing possibility. It is of interest to note, however, that similar to meditation studies, no structural differences were found for BA 46, which is often also considered a part of the DLPFC.
In review of the evidence above it appears that both lucid dreamers and meditators are known for heightened levels of dreaming and waking metacognitive awareness, respectively, which are related to structural differences in the DLPFC (BA
9) and FPC (BA 10). Given that this literature review departs from the viewpoint that metacognitive awareness is a cognitive process relatively preserved across the sleep wake cycle, it should also be apparent that meditators are also more likely to be lucid dreamers. Indeed, there is supporting evidence for this. A relationship between meditation practice and the occurrence of lucid dreams has been found multiple times (Gackenbach, 1978, 1981, 1990, cited in Schredl & Erlacher, 2004; Gackenbach, Cranson, & Alexander, 1986, cited in Schredl & Erlacher, 2004; Hunt, 1991, cited in Schredl & Erlacher, 2004; Levitan, 1993, cited in Schredl & Erlacher, 2004). Meditators are more prone to have lucid dreams. Furthermore, a population study by Stumbrys, Erlacher, and Malinowski (2015) found a direct relationship between metacognitive awareness during the day and night in meditators. 528 participants (290 men, 238 women) recruited through a lucid dreaming website responded to an online questionnaire regarding dreaming, meditation, and mindfulness. The measurement instrument for assessing mindfulness was the Freiburg Mindfulness Inventory (FMI; Walach, Buchheld, Buttenmuller, Kleinknecht, & Schmidt, 2004; cited in Stumbrys et al., 2015), which consists of two factors (acceptance and presence). Acceptance refers to the accepting and appreciative attitude towards experience, while presence refers to sustaining full awareness of experience as it is happening. As would be expected, Stumbrys et al. (2015) also found supporting evidence for the notion that individuals who reported having prior meditation experience also reported higher lucid dream frequency (4.28 lucid dreams a month as compared to 2.55 lucid dreams a month for non-meditators). Concerning the direct relationship between wake and sleep metacognitive awareness, there was a positive correlation found the between FMI mindfulness score and lucid dreaming frequency in meditators. Furthermore, lucid dream frequency was more strongly associated with mindful presence rather than
acceptance. Thus individual mindfulness is positively related to lucid dream frequency in meditators and is most strongly related to the facet of mindful presence (being able to sustain full awareness to the experience while it is happening). The authors concluded that higher awareness cultivated during daytime is also reflected in higher awareness of one`s mental states while dreaming. This relationship between meditation and lucid dreaming provides additional supporting evidence that there is indeed the same core cognitive mechanism with the underlying neural substrates of BA 9 and 10 that supports greater metacognitive awareness in both the wake and sleep states.
To summarize, it appears that trait-level metacognitive awareness can be maintained throughout wake and REM sleep. The ability to lucid dream can be trained from within the waking state, lucid dreamers perform better on a variety of tasks requiring involvement of the DLPFC and report greater control over their waking life, the latter also likely being dependent on metacognitive awareness (Purcell et al., 1986; Tholey, 1983; Blagrove et al., 2010; Bourke & Shaw, 2014; Neider et al., 2011; MacDonald et al., 2000; Qiu et al., 2010; Brevers, et al., 2013; Blagrove & Tucker, 1994; Blagrove & Hartnell, 2000; Gruber et al., 1995; Farb et al., 2010). Lucid dreamers also show greater GMV in the DLPFC and FPC, brain regions implicated in metacognitive awareness (Filevich et al., 2015). In addition, meditators, a population known for greater than average metacognitive awareness and greater GMV in the DLPFC and FPC, are more likely to be lucid dreamers (Gackenbach, 1978, 1981, 1990, cited in Schredl & Erlacher, 2004; Gackenbach, Cranson, & Alexander, 1986, cited in Schredl & Erlacher, 2004; Hunt, 1991, cited in Schredl & Erlacher, 2004; Levitan, 1993, cited in Schredl & Erlacher, 2004). Furthermore, the level of day-time metacognitive awareness in meditators was also positively related to lucid dreaming
occurrence (Stumbrys et al., 2015). This provides direct supporting evidence that it is indeed the level of day-time metacognitive awareness that relates to lucid dreaming.
These studies (Purcell et al., 1986; Tholey, 1983; Blagrove et al., 2010; Bourke & Shaw, 2014; Neider et al., 2011; MacDonald et al., 2000; Qiu et al., 2010; Brevers, et al., 2013; Blagrove & Tucker, 1994; Blagrove & Hartnell, 2000; Gruber et al., 1995; Farb et al., 2010; Filevich et al., 2015; Gackenbach, 1978, 1981, 1990, cited in Schredl & Erlacher, 2004; Gackenbach, Cranson, & Alexander, 1986, cited in Schredl & Erlacher, 2004; Hunt, 1991, cited in Schredl & Erlacher, 2004; Levitan, 1993, cited in Schredl & Erlacher, 2004; Stumbrys et al., 2015) offer opposing evidence for the hypothesis presented by Voss et al. (2009; 2014), which states that that lucid dreaming is a dissociated state containing features of both wake and dreaming (intermediate/hybrid state between wake and sleep), i.e. a higher level of metacognitive awareness, which is indicative of the waking state, is maintained when falling asleep. Instead, it offers supporting evidence for the alternative viewpoint presented in this literature review, which states that lucid dreaming may actually entail a general trait-level of greater metacognitive awareness (differential functioning), which is preserved across the sleep and wake cycle (i.e., lucid dreamers were strictly more metacognitively aware to begin with and thus maintain this awareness for REM dreams as well.)
However, as per definition, lucid dreams occur exclusively during non-lucid REM sleep when metacognitive awareness is achieved when one realizes that he or she is dreaming (LaBerge, 1981; Payne, 2014). In other words, it is commonly believed that lucidity can only arise out of non-lucid dreams. This does imply that there was a gap in metacognitive awareness before it was regained. Even though potentially inconceivable for the status quo of lucid dreaming and metacognitive
awareness research, it may be that certain populations may have such highly developed levels of metacognitive awareness in that it could be maintained throughout the day and night, including for NREM sleep (Norbu & Katz, 1992; Wangyal, 1998). Even though research on this is limited, there is still some evidence suggesting that this is possible.
NREM Sleep Metacognitive Awareness: The Missing Link
Given that there is evidence supporting a relationship between REM sleep and wake time metacognitive awareness, it is of interest to investigate if metacognitive awareness can also be maintained during NREM sleep. Lucid NREM dreaming is alleged to be possible by employing a variety of methods to retain lucidity while falling asleep and entering NREM sleep. As stated by Tholey (1983), “in these methods one concentrates on visual phenomena, one’s body or one’s own thinking ego while falling asleep” (p. 79). Research on NREM lucidity, however, is very limited and consists of primarily three studies (Dane & Van De Castle, 1984; Laberge et al., 1981; Stumbrys & Erlacher, 2012). LaBerge et al. (1981) described one subject who reported signaling of lucidity during descending N1, however EOG signaling was not found, leading to no scientific proof for lucid NREM sleep. Dane & Van De Castle (1984) utilized posthypnotic suggestions in a hypnotically susceptible population to prime subjects to signal lucidity during NREM sleep and obtained an unusually high number (17) of NREM lucidity signaling. However, given the use of hypnosis as an induction method, this study did not investigate trait-level NREM lucidity. Stumbrys & Erlacher (2012) were the first to find incidences of lucidity signaling during N2 in two adept lucid dreamers while conducting a study on REM lucid dreaming without the use of lucidity induction methods. These findings, albeit limited, open up the possibility that lucidity is not only possible for REM sleep, but also for NREM sleep. In addition, the phenomenological description of one subject of Stumbrys & Erlacher (2012) was rather different from lucid REM mentation and non-lucid REM mentation common in sleep mentation awakening studies. This subject reported that there was no visual imagery present, but there was a floating sensation
without feeling her body, therefore she realized that it has to be a dream and gave the eye-signal.
Concerning the mentation in these lucid sleep states; with REM sleep lucidity it can be assumed that the mentation is similar to wake state mentation, with the exception that dream stimuli are internally generated, as opposed to externally perceived, likely the reason for requiring additional activation of the FPC next to the DLPFC for monitoring of this internally generated information (Christoff & Gabrielli, 2000). REM is indeed often referred to as paradoxical sleep due to the presence of wake-like low amplitude high frequency EEG (Peigneux et al., 2001).
NREM brain activity, however, is characterized by higher amplitude and slower frequency activity, in particular for deep sleep (N3) (Medicine & Iber, 2007). In accord, reported mentation is also very different. Classical sleep mentation research findings pin NREM states as thought-like, ruminative and free of elaborate visual stimuli (Fosse et al., 2004; Hobson & Stickgold, 1994; Kales et al., 1967). Especially when awoken from deep sleep (N3) and given the instruction to report anything that was going through their minds just before waking up, people tend to report short and fragmentary thoughts or are unable to report anything at all (Tononi & Koch, 2008).
According to Thompson (2015), however, the current taxonomy of NREM sleep mentation can be refined. Indeed, it may be that the findings obtained in the modern West characterizes non-lucid NREM by this ruminative minimally conscious mentation. This would be in accord with the continuity hypothesis that if a great portion of the waking state is spent mind-wandering and ruminating, for a similar process to continue during the sleeping state (Domhoff, 1996). Nonetheless, in subjects practicing advanced forms of meditation or lucid NREM sleep, the goal is to attain states of Non-Dual Awareness (Josipovic, 2010; L. I. Mason & Orme-Johnson,
2010). Both the more commonly known focused-attention and open-monitoring styles of meditation contain an essentially dualistic orientation of ‘subject-observing-object’ (Lutz et al., 2008). There is, however, an additional style of meditation that does not employ this strategy, but instead relies on accessing a level of awareness that is inherently free from this dualistic subject–object construct. Non-Dual awareness has also been termed open awareness or open presence (Kozhevnikov, Louchakova, Josipovic, & Motes, 2009; Lutz, Dunne, & Davidson, 2007). Thompson (2015) illustrates Non-Dual Awareness with a single anecdotal lucid sleep report:
I am suspended in space—dream space, I think. There is nothing here, just millions of greyish dots and I am one of the dots, there’s no dream-body anymore, I’m just a dot [of] pure consciousness suspended. A feeling of great peace comes over me and a sense of gentle, infinite expansion. It’s as if everything and nothing are the same thing and there is a sense of effortless belonging. As the sense of expansion increases I am no longer a single dot of consciousness; all the dots are me and I am them. There’s no “I” or “them.” We are one. There’s just a blissful sense of timelessness and oneness and a merging with the light. After an indefinable length of time, I start to feel the weight of my body in bed, and settle back into it, tingling all over. (Clare Johnson, unpublished dream report, March 19, 1995; cited in Thompson, 2015)
As is evident with the report of Stumbrys & Erlacher (2012) and the descriptive report of Thompson (2015), it appears that a lack of bodily sensation, a feeling of being suspended in space and no elaborate visual dream mentation may be common for non-REM lucidity. Consequently, lucid NREM research may also come with a shift of understanding over the quality of primary consciousness of which subjects are metacognitively aware of.
Even though not directly under the research area of NREM lucidity, additional research into NREM metacognitive awareness comes from a related research field on “witnessing” sleep. According to Mason & Orm-Johnson (2010) the development of metacognitive/self awareness during sleep does not end with lucidity. While with lucid dreaming, one is actively interacting with the dream content, one can move further along the continuum to a quieter, uninvolved state of awareness known as witnessing. To better illustrate the notion of witnessing with a movie analogy provided by Travis, Arenander, and DuBois (2004), non-lucid mentation is characterized by identifying with the movie and “losing” one’s self in it, lucid mentation is characterized by realizing it is a movie but still going along with and being captivated by the storyline while with witnessing an additional step back is taken and the most salient part of the experience becomes self-reflective awareness. Thus, the most salient facet becomes the experience of (metacognitive) awareness itself, as opposed to the content, while the reverse could be stated for non-lucidity.
There have been accounts of advanced transcendental meditation (TM) practitioners reporting witnessing deep sleep (Banquet & Sailhan, 1974; L. Mason, Alexander, Travis, & Gackenbach, 1990; L. I. Mason et al., 1997). Mason et al. (1997) measured 11 TM practitioners with on average 17.8 years of practice as compared to a control group consisting of novice-practitioners and a control group consisting of non-practitioners using a 4-channel EEG setup. The EEG tracings of the practitioners included simultaneous 7-9 Hz with delta activity during deep sleep (N3) (Mason et al., 1997). Spectral analysis revealed increased alpha/theta (7-9 Hz.) relative power during N3 in practitioners as compared to the controls with no difference in delta power. In addition, there was a graded difference in alpha/theta power during deep sleep, with long-term practitioners having greater alpha/theta
power than short-term practitioners, who in turn had greater alpha/theta power than non-practitioners. This implied a causal relationship of years of practice on alpha/theta power. These findings were consistent with earlier pilot findings of simultaneous alpha with delta during deep sleep in sleep of 12 TM practitioners versus controls (Banquet and Sailhan, 1974; Mason et al., 1990). One subject could even signal witnessing (moving eyes in a pre-agreed manner during sleep to signal witnessing) during delta waves indicating extended awareness and ability to signal (Banquet & Sailhan, 1974).
However, methodologically speaking, there are a few limitations with these studies on witness consciousness during sleep given that they are primarily based on self-reports and only one subject could signal witnessing. In addition, given that these are relatively old studies, low-density EEG recordings were obtained and source localization was not attempted. Furthermore, these studies rely quite heavily on the Transcendental Meditation (TM) research framework and terminology rather than a cognitive neuroscience framework (e.g. terms as enlightenment and pure awareness to describe the phenomenonology of these states are commonly used).
What can be deduced from these studies is that a more developed form of metacognitive awareness may be needed to maintain metacognitive awareness throughout NREM sleep. NREM sleep lucidity may be difficult to grasp given the lack of elaborate mentation, which is usually present in the wake and REM states. This may be related to axioms on which neuroscience currently departs from. The current neuroscientific approach in greater understanding of the mind is based on Cognitivism, a stream of Psychology, which originated in the 1950’s that departs from an information processing approach (Huitt, 2003; Still & Costall, 1991). Cognitivism likens the human mind to a computer in that they are both believed to be adept content
driven information processors. For example, Tononi (2008) equates consciousness to integrated information processing and holds a staunch view of the disappearance of consciousness during deep sleep due to it being characterized by limited information processing. It may indeed be difficult to fathom that a highly adept form of awareness can be maintained during such a state of minimal information processing (Huitt, 2003; Thompson, 2015). Given the lack of content driven information processing, awareness itself would inadvertently need to take the foreground of the focal point of attention. This might indeed require an additional step back in metacognitive awareness to a state of “witnessing”. The witnessing state indeed entails more attention towards the phenomenology of experience itself as opposed to its content.
In conclusion, even though not thoroughly investigated, there is some supporting evidence for metacognitive awareness during NREM sleep, albeit describing a rather altered state of consciousness (Dane & Van De Castle, 1984; LaBerge, 1981; Stumbrys & Erlacher, 2012; Thompson, 2015; Mason et al., 1997; Mason et al., 1990; Banquet & Sailhan, 1974). Sporadic reports on NREM lucidity have been apparent since the beginnings of lucid dreaming research (LaBerge, 1981; Dane & Van De Castle, 1984). Stumbrys et al. (2012) were the first to obtain measurements of signaled lucidity in adept lucid dreamers during NREM sleep. A related field of meditation research on witnessing sleep has also obtained evidence for the maintenance of metacognitive awareness during NREM sleep (Mason et al., 1997; Mason et al., 1990; Banquet & Sailhan, 1974. These studies provide, albeit limited, supporting evidence for the notion of the ability to maintain metacognitive awareness through all parts of the sleep/wake cycle. More research, however, is required on NREM lucid dreaming for greater understanding of this phenomenon.
Conclusion & Discussion
This literature thesis reviewed whether metacognitive awareness can be maintained across the sleep/wake cycle. More specifically, it was assessed to what extent the commonly believed notion that sleep is universally characterized by a loss of metacognitive awareness as compared to the waking state can be upheld. It is evident that metacognitive awareness, which is subserved by the FPC and DLPFC, is a much more intermittent process during the wake state in the normal population than is often believed (Christoff & Gabrielli; Schooler et al., 2011). Individuals are believed to reside in a state mind-wandering, which is characterized by a lack of metacognitive awareness, up to 50% of their time spent awake (Killingsworth & Gilbert, 2010). There are, however, also trait-level differences when it comes to metacognitive awareness in that mindfulness and meditation practices are thought to increase it during the waking state (Austin, 1998; Ortner, Kilner, & Zelazo, 2007). Furthermore, it is also evident that sleep does not need to be characterized by a loss of metacognitive awareness. Lucid dreaming serves as a well-researched model of metacognitively aware REM sleep (Van Eeden, 1913; LaBerge, 1981). Neuroimaging studies also indicate that a relatively preserved level of prefrontal activity, in particular the DLPFC and FPC, characterizes REM lucid dreaming (Dresler et al., 2012; Voss et al., 2009). In addition, experimentally manipulating the DLPFC activation leads to an increase of metacognitive awareness during REM sleep (Stumbrys et al., 2013; Voss et al., 2014). The individual differences present for both wake and sleep metacognitive awareness points to the intriguing possibility that wake and sleep levels of metacognitive awareness may be related (Austin, 1998; Ortner, Kilner, & Zelazo, 2007; Erlacher, Schredl, Watanabe, Yamana, & Gantzert, 2008; Erlacher, Stumbrys & Schredl, 2011). Indeed, mind-wandering and non-lucid REM
mentation are very similar in first-person content and brain activity (Fox et al., 2012). Also, studies investigating the similarities between sleep and wake mentation found subjects maintained a much more comparable level of metacognitive awareness to waking state mentation during dreams than what is commonly believed (Kahan, & Laberge, 1996).
Furthermore, there appears to be a direct relationship between high levels of waking state metacognitive awareness and the ability to lucid dream. Meditators, a population known for heightened waking state metacognitive awareness, are more likely to be lucid dreamers; lucid dreamers perform better than the normal population on a variety of complex cognitive tasks that require DLPFC activation and also report feeling more in control of their waking state (Gackenbach, 1978, 1981, 1990, cited in Schredl & Erlacher, 2004; Gackenbach, Cranson, & Alexander, 1986, cited in Schredl & Erlacher, 2004; Hunt, 1991, cited in Schredl & Erlacher, 2004; Levitan, 1993, cited in Schredl & Erlacher, 2004; Stumbrys et al., 2015; Blagrove et al., 2010; Bourke & Shaw, 2014; Neider et al., 2011; MacDonald et al., 2000; Qiu et al., 2010; Brevers, et al., 2013; Blagrove & Tucker, 1994; Blagrove & Hartnell, 2000; Gruber et al., 1995). Lucid dreamers also have greater GMV in regions implied in metacognitive awareness (Filevich et al., 2015). In addition, there is also evidence for metacognitive awareness during NREM sleep, which supports the notion that metacognitive awareness can be maintained throughout the sleep/wake cycle (LaBerge, 1981; Dane & Van De Castle, 1984; Stumbrys et al., 2012; Mason et al., 1997; Mason et al., 1990; Banquet & Sailhan, 1974).
From the discussed literature it becomes evident that lucid dreaming is the manifestation of a trait of greater metacognitive awareness preserved across the sleep/wake cycle, rather than the commonly believed viewpoint that lucid dreaming is
