The Science of Lucid Dreaming

Introduction

Distinction between dream and waking reality can be a serious problem: the most striking example is that most of the dreams are recognized as reality, until we wake up and realize that we were dreaming. In some cases, however, the dreamer is able to realize that he or she is dreaming, while being still in a dream state. Becoming aware of the fact that everything what surrounds us is a product of dream imagery can lead to an extraordinary experience, a state of mind between waking and dreaming, the so-called lucid dreaming.

Lucid dreams, i.e. dreams during which the dreamer is aware of the fact that he or she is dreaming, has been known for centuries. 350 years B.C. in his classic work “On Dreams” Aristotle noted that “for often, when one is asleep, there is something in consciousness which declares that what then presents itself is but a dream”, which was the first known description of lucid dreaming. For Tibetan Buddhists lucid dreaming for more than a thousand years served as a part of a practice leading to enlightenment and the discovery of the ultimate nature of the self (LaBerge and Rheingold 1990). In modern times, lucid dreaming has been popularized by Marquis Leon Hervey de Saint-Denys. In his book “Dreams and How to Guide Them”, the author used for the first time the French term “rêve lucide” which refers to “lucid dream” (Hervey de Saint-Denys 1867). The phrase “lucid dream” has been coined by Frederik van Eeden in 1913 in his scientific paper “A Study of Dreams”, which was the first extensive scientific report concerning lucid dreaming written in English (van Eeden 1913). For a long time lucid dreaming has been disregarded by most of dream researchers, until late 1970s, when a new approach based on communication between subject (dreamer) and experimenter had been applied for the first time.

Although during sleep general muscle tonus is inhibited, electrical activity of limb muscles can be measured and correlation between dreamed limb movements and EMG (electromyogram) activity in the corresponding limb can be observed. Also the dreamed gaze shifts are related to the EOG (electrooculogram) pattern recorded during sleep. Therefore, some specific dream actions, like hand clenching or eye movements that would be seen on polygraph, could serve as a behavioral response signaling the onset of a lucid dream (LaBerge et al. 1981, LaBerge 1990). Beside the verification of the occurrence of lucid dreaming, the fact that trained lucid dreamers can remember pre-sleep instructions and signal it to the researcher in laboratory allows to carry out different dream experiments. For example, by marking the beginning and the end of specific tasks which lucid dreamers are instructed before the night to perform during lucid dream, some psychophysiological (e.g. heart or respiration rate) and electrophysiological (brain electric activity) correlates of performed tasks can be measured (LaBerge et al. 1981, LaBerge 1990, Erlacher and Schredl 2008). This valuable finding allows the methodical testing of hypotheses, which was not accessible earlier in lucid dream research.

How is the dream content shaped?

With the discovery of the rapid eye movement (REM) sleep (Aserinsky and Kleitman 1957) it has been proposed that dreaming is exclusively related to that phenomenon. It has also been suggested that there is a causal link between eye movements in REM sleep and the dream content. The so-called “screening hypothesis” assumed that rapid eye movements are saccadic scans of objects present in the dream scene [This issue, however, remains a question; there are studies which demonstrate this link, and others that present contradictory data (for a review see Muzur 2005)]. However, during the experiments in which subjects were asked to recall the dream content after awakening in different stages of non-rapid eye movement (NREM) sleep (e.g. Kales et al. 1967), they reported dream experiencing, but less remarkable in quantity, vividness, bizarreness and emotion, when compared to REM dreaming. For a long time many studies have failed to answer the question whether NREM dreaming originates from the NREM sleep mechanisms. It was suggested, among others, that NREM recall reports are containing experiences from the preceding REM sleep or are actually a stage 2 hypnagogic hallucinations [i.e. hallucinations occurring at sleep onset. Hallucinations which may occur upon awakening are called hypnopompic] (Muzur 2005). Nielsen (2000) proposed that NREM dreaming is a result of “covert” REM sleep, which means that some of REM sleep processes can dissociate from REM sleep and stimulate NREM mentation (Nielsen 2000). Some studies, however, indicated that dreaming during non-rapid eye movement sleep can occur in the absence of prior rapid eye movement sleep (Suzuki et al. 2004).

REM sleep contains phasic and tonic periods. Bursts of rapid eye movements occur only in phasic periods of REM sleep and are absent during tonic REM sleep (Moruzzi 1963). Characteristic features of phasic REM sleep are pontine – geniculate – occipital spikes (PGO waves) (Callaway et al. 1987). There is a body of evidence supporting an important role of PGO waves in many different neural processes, such as network organization in REM sleep, brainstem activation, or the transmission of eye movement information to the cortex (cf. Lim et al. 2007). Although the existence of PGO waves in humans has not been definitely proven, there is some indirect evidence (based on both electrophysiological and neuroimaging data), which supports this hypothesis (cf. Muzur 2005). Lim et al. (2007) demonstrated the occurrence of REM sleep associated with phasic activity in the human pons, which was coupled with characteristic cortical potentials, and concluded that PGO waves are features of human REM sleep. According to Lim et al. human PGO waves are generated or propagated in a region of the pontomesencephalic tegmentum, however they are not fully associated with eye movements, and are associated with changes in cortical activity (Lim et al. 2007).

It is also proposed that PGO waves may also contribute to dream generation. According to activation-synthesis dream hypothesis (Hobson and McCarley 1977, Hobson et al. 2000) sensorimotor and limbic regions of the forebrain produce a coherent experience from the incomplete and random inputs received from the brain stem. The shift in input source, from the formed visual images on retina in waking to the chaotic brain stem stimulation of REM sleep occurs in the context of quiescence of the brain stem noradrenergic and serotonergic neurons. As a consequence of aminergic disinhibition, hyperexcitability of cholinoceptive peribrachial neurons occurs and leads to phasic activation of the lateral geniculate bodies and visual cortex. Although the brain stem and visual system are deafferentiated, cholinergically stimulated and aminergically demodulated, the brain stem signals still convey the information about the direction of saccadic eye movements to the forebrain (cf. Hobson et al. 2000). Therefore, dreaming results as an interpretation of information concerning rapid eye movements and activated (phasic) brain stem motor pattern generator by a cortex. It has been proposed that cortical and limbic regions, when cholinergically activated by REM sleep events such as PGO waves, can synthesize their own information. According to this theory, dream hallucinations must incorporate also a visual material from a variety of memory sources (Hobson et al. 2000). It seems, however, that the dream content does not accurately correspond to episodic memory due to lack of episodic memory replay during dreaming (Fosse et al. 2003), but rather isolated, incomplete fragments of narrative memory are synthesized into a new dream scenario. Recalling an integrated episodic memory during waking involves hippocampal outflow of the information to the different sites of the cortex. Moreover, access to episodic memory depends on control systems localized in the dorsolateral prefrontal cortex (DLPFC) (for a review see Simons and Spiers 2003). Because DLPFC is deactivated during NREM and REM sleep (Maquet et al. 1996, Nofzinger et al. 1997, Brown et al. 1997), limited contribution of episodic memory to a dream formation seems to be possible. [In fact, the “deactivation” of prefrontal cortex during sleep should be considered in terms of relative activity (Muzur et al. 2002). In fact, DLPFC deactivation represents the inhibition of inadequate element combining, which is responsible for constant monitoring of morally or logically inadequate contents in waking (Muzur 2005)].

Changes in sensory input source and neuromodulation during sleep probably result in such cognitive features of dreaming as the hallucinatory visual imagery, frequent attention shifts, emotional intensification of experienced fillings (linked with selective activation of subcortical and cortical limbic structures in REM sleep), loss of voluntary control and memory loss within and after dreaming (Hobson et al. 2000). As a consequence of deactivation of DLPFC during sleep executive functions, which are attributed to this brain region, are attenuated and, therefore executive features such as self-consciousness or analytical and conceptual thinking are severely impaired during NREM sleep and are highly decreased in REM sleep (Muzur et al. 2002). 

Consciousness during dream

During sleep consciousness undergoes profound alteration (Dietrich 2003). The most robust and universal change of consciousness associated with transition from waking to sleep is the loss of self-conscious awareness (Muzur et al. 2002). According to Posner and Rothbart (1998), the problem of consciousness includes both volition and awareness. REM sleep is a specific condition in which those two features are dissociated: the voluntary control is lost but awareness can remain at a high level. Rechtschaffen (1978) divided waking consciousness into two prevalent streams: voluntary control and reflective awareness. The reflective stream of awareness is highly attenuated in dreams, which makes us unable to discriminate between dream and reality. Second stream, volition, is also absent, since volitional control, at least phenomenologically, implies reflectiveness. Some researchers, however, argued that reflective awareness and other metacognitive experiences are much more common in dreams than previously thought, because they are often underreported in narrative dream reports (Kahan 1994, Kahan and LaBerge 1994, Kahan et al. 1997).
 
Consciousness as awareness (phenomenological meaning) can be distinguished into three types. First type of awareness, the primary phenomenal experience of objects or events (cf. primary awareness) usually occurs during dreaming (Cicogna and Bosinelli 2001), at least in the form of a dream content (Posner and Rothbart 1998). According to Cicogna and Bosinelli (2001) the dreamer typically is aware of events and actions which take place in the dream, and that he participates in it, but also is aware of being himself (self-awareness), though many modifications are possible (e.g. awareness can be preserved only in respect to the feeling oneself expressed by the identification with other characters or objects, or as a representation of oneself as a passive observer of the dream scene). The third type of awareness, meta-awareness, is considered to be an awareness of having an awareness of the first type or, in other words, the awareness of having own thoughts and behavior. This type of awareness is usually not present during dreams, however, if meta-awareness is fully preserved then the dreamer can realize that he is dreaming, while the dream is in progress (Cicogna and Bosinelli 2001), which may result in lucid dreaming.


 Brain EEG activity during lucid dreaming

Since lucid dreaming has been verified under laboratory conditions, relatively few studies investigating the brain mechanisms of this phenomenon have been done so far. Early research showed that lucid dreams usually occur in phasic REM sleep, which is associated with high autonomic nervous system activity such as increased and irregular respiration rate and bursts of rapid eye movements. These findings indicated that periods of relatively high brain activation are needed for consciousness to be attained during dreaming (LaBerge et al. 1981, LaBerge et al. 1986). However, the issue whether the brain activation was general or local remained unanswered.

First suggestion concerning brain areas which may be involved in lucid dreaming came from early EEG analysis performed by LaBerge and colleagues at Stanford University (for a review see LaBerge 1993). A study of the left/right ratios of alpha activity during lucid dreaming indicated that the onset of lucidity is correlated with a decrease of alpha activity in the left parietal region. Decreased alpha activity over specific brain area is generally considered as an activation index of this cortical region, both in active wakefulness (Schupp et al. 1994) and mental imagery (Kaufman et al. 1990). This finding is consistent with a study investigating the distribution of brainwave activity during the periods before and after the onset of consciousness during a lucid dream. Among different frequency bands of EEG analyzed in this study, the most interesting results concerned the alpha band (8-12 Hz), where the activation of posterior left hemisphere (decreased alpha activity over this area) was recorded during the first 30 seconds of lucidity (LaBerge 1993).

Increased activity over the left parietal area was also observed in a more recent study performed by Holzinger et al. (2006). The aim of this research was to investigate the existence of physiological differences between lucid and non-lucid REM sleep epochs. During the epochs of lucid dreaming the greatest increase of beta-1 (13-19 Hz) activity was found in parietal regions. The highest increase was observed in the left parietal lobe, however, in this case the level of statistical significance was not attained. Holzinger et al. (2006) also found an interaction between front/back activity and lucidity. During lucid epochs an overall decrease in beta-1 activity was observed in frontal areas, whereas in parietal regions an overall increase was recorded. The ratio of frontal to parietal activity in lucid dream epochs was approximately 1 to 1.77, while in non-lucid epochs the ratio obtained 1 to 1.16.

Holzinger et al. (2006) suggest that the activation of the left parietal region during lucid dreaming is due to the fact that this region is considered to be engaged in semantic understanding of the meaning of words such as “This is a dream”. Also LaBerge (1993) stated that lucid dreaming should be associated with the activation of the left hemisphere, where the language center is localized, since to become aware of the fact that one is dreaming it is required to spell out to oneself: “This is a dream” (LaBerge 1993).

Holzinger et al. (2006) also refer to Taylor (1990) who describes the inferior parietal lobule (IPL) as an essential brain region for consciousness, however, they do not provide any further explanation of this link. Indeed, Taylor’s theory of the central representation considers inferior parietal lobule as a brain region which may be implied in consciousness because of following reasons. Firstly, Taylor indicates that there is an evidence for the essential nature of the IPL in working memory (cf. LaBerge 1990), particularly for spatial locations and for objects. Secondly, both superior parietal lobule (SPL) and IPL were found by Hirsh et al. (1999) to be active during a functional magnetic resonance imaging (fMRI) study, in which a subject was asked to imagine rotation of self in space. Thirdly, disturbed function of this region, due to lesions, or as a consequence of schizophrenia, results e.g., in impaired perception of objects in the external world (Taylor 2001). Taylor emphasizes the importance of the IPL “in the creation of consciousness of objects in the external world or held in imagination through attention and working memory as well as in having important information from body sites as to body positions or actions being taken”. Moreover, if information from the body has a direct influence on the conscious experience, this implies that the IPL is well suited to carry the so called “Central Representation”, defined as follows: “The Central Representation (CR) is the combined set of multimodal activations involved in fusing sensory activity, body positions, salience, and intentionality for future planning: it involves a competitive process between the various modules it contains to single one out to be the content of consciousness, with information bound to it usable for report to other working memory sites for further planning or action” (Taylor 2001, p.401-402).





Another, more recent study, conducted by Voss et al (2009) showed that lucid dreaming is not only a part of REM sleep but also has some features of waking state. During lucid dreaming, as well as in REM sleep, increased power of delta and theta frequency bands were observed, accompanied with lower alpha power. Because elevated alpha power is characteristic of of the state of waking with eyes closed, this reversed pattern is an evidence that lucid dreaming occurs in a state of sleep. However, the authors also found  wake-like activation higher frequency band (~40Hz, gamma) during lucid dreaming compared to normal REM sleep, which occurred frontolaterally and frontally. In fact, increased gamma activity in lucid dreaming was already observed by Mota-Rolim et al (2008) in left prefrontal lobe, together with significantly more alpha activity in the right posterior temporal lobe, compared to non-lucid REM sleep.

Different activities, both cognitive tasks and motor performances, executed during lucid dreaming, seem to share the same brain mechanisms and involve the same cortical areas as in waking. In a study by Erlacher et al. (2003) a single lucid dreamer was instructed to execute either left or right hand clenching, and each dream event was marked with pre-arranged eye-movement patterns. This allowed analyzing the EEG alpha power over the bilateral motor areas (C3, Cz, C4) for each motor performance. Results showed that, as in waking, executed hand clenching was associated with a decrease of alpha activity over the motor cortex. During dreamed counting (control condition) the activation of motor areas was not observed.Similar results were obtained recently using functional magnetic resonance imaging (fMRI) technology. [For further description, see below].

Recently, Strelen (2006) investigated the event-related evoked potentials (P300) during lucid dreaming using the odd-ball paradigm with acoustic modality [One of typical procedure of evoking the P300 wave in which the target acoustic stimulus is presented amongst more frequent background non-target stimuli]. Before the night the participants were listening to two types of short acoustic signals, high and low tone. Then, they were instructed to signal by a single left-right eye movement each time they hear the high tone (the target stimulus) during lucid dream. In all analyzed participants (3 subjects) EEG analysis revealed P300 activation correlated with EOG recordings of the eye movement signaling pattern. Thus, P300 activation was in fact correlated with the identification of the target stimulus (high tone) Therefore, the appearance of P300 wave can be interpreted as conscious processing of acoustic information. Moreover, such EEG pattern is characteristic for wakefulness.(Description of this experiment is taken from Erlacher and Schredl 2008).

An interesting study demonstrating similarities of brain functioning between waking and lucid dreaming was carried out by LaBerge and Dement (1982). The authors assumed that performed cognitive tasks, singing and counting, would be associated with the activation of left and right cerebral hemispheres, respectively. The idea was based on numerous reports of cognitive tasks dependency of EEG alpha activity lateralization in waking state. The experienced lucid dreamers when they became lucid, executed pre-arranged eye-movement pattern, sang for ten seconds, signaled again to mark the task changing moment, counted for ten seconds, and again signaled to mark the end of the task. The EEG alpha activity was derived from electrodes placed over the left and right temporal lobe (T3/Cz and T4/Cz). The results supported the hypothesis of lateralization of alpha activity during performance of different cognitive tasks: similarly as in waking, the right hemisphere showed less alpha activity during singing, but during counting decrease of alpha activity was observed in left hemisphere. In control condition, where the same tasks were imagined, no significant laterality of EEG alpha activity shifts was observed. These results, however, were obtained in a small group of only four subjects. Thus, more evidence should be provided to hold these conclusions (LaBerge and Dement 1982).

Presence of alpha activity during REM sleep has been usually considered as a sign of relative cortical activation and wakefulness, however, findings obtained by LaBerge and Dement (1982) suggested quite the contrary. As authors noted, when a person awakes from a vivid dream to a dark room, alpha power increases, at least over occipital area. Alpha activity suppression over occipital region (visual cortex) during phasic-REM periods may be considered as an electrophysiological index of the visual imagery presence during oneiric experience (Cantero et al. 1999). REM alpha activity has been also proved to successfully discriminate between pre-lucid, lucid and non-lucid dreams (Ogilvie et al. 1982, Tyson et al. 1984). Highly emotional, bizarre and irrational dreams may lead to a pre-lucid dream, in which the dreamer develops a critical attitude toward the dream. This in turn may lead to questioning the reality of a dream and recognizing it as a dream and not reality. The highest alpha activity has been associated with pre-lucid dreams and highly bizarre content, whereas lucid dreams were found to have high alpha activity early in the REM period followed by a distinct lowering of REM alpha. This supports the hypothesis that consciousness during dream can sometimes emerge from pre-lucid experience (Tyson et al. 1984). However, when Ogilvie et al. (1988) attempted to test this hypothesis based on an EEG data obtained from a high-frequency lucid dreamer, they did not find any important differences in EEG alpha activity. Instead, they found differences in the percent rate of theta activity, which was higher during REM lucid dreaming than during stage 2 lucid dreaming (Ogilvie et al. 1988).

Recently, there have been made some attempts to use neuroimaging techniques, like fMRI,  in order to determine the changes in brain activity during lucid dreaming. [For overview of this experiments, see below].
 

Cognitive Psychology Studies

Another study supporting the hypothesis of brain lateralization of cognitive functions during lucid dreaming was conducted recently by Robert Piller (2009). Based on the postulate that lucid dreaming could be characterized by greater right hemisphere activation (Green and McCreery 1994), Piller attempted to test the right hemisphere prevalence hypothesis during lucid dreaming by means of performance of different cognitive tasks attributed to the left (LH task) or right (RH task) hemisphere. Participants were instructed to rate the difficulty of performing each activity in a lucid dream. Tasks were divided into three groups. Each group contained a pair of LH and RH tasks: reading a sentence and observing a painting, writing a sentence and drawing a cube, and saying a sentence and humming a tune. The main assumption was that, if the right hemisphere is specialized in lucid dreaming, then LH tasks should be more difficult to perform in a lucid dream than RH tasks. In the first group, according to expectations, reading was harder to perform than observing, especially for right-handed subjects (due to more pronounced cerebral asymmetry). In the second group of tasks drawing was overall more difficult than writing while the subject was lucid dreaming, which is inconsistent with the right hemisphere hypothesis, and in third group LH task (speaking) tended to be harder to perform than RH task (humming), though the level of statistical significance was not reached. The author explains the unexpected result obtained in a second task group by the fact that left hemisphere is “more detail-oriented” and is better in controlling sequential hand movements, so, ultimately, both speaking and humming are actually LH tasks. However, the hypothesis proposing the lateralization of cognitive functions during lucid dreaming is supported by the fact that all three groups of tasks differed between left- and right-handed participants (Piller 2009). 

Cortical activation and working memory

Electrophysiological studies on lucid dreaming suggest that an elevated level of neuronal activation is the necessary condition for consciousness to emerge during a dream. LaBerge (1990) advocated the view that the high level of CNS activation is evidently linked to a high level of cognitive function involved in lucid dreaming, and that becoming lucid requires sufficient level of working memory which is necessary to recognize the dream as a dream.

Working memory plays a pivotal role in the organization of complex cognitive functions such as planning, conceptual thinking, problem solving and decision making. Working memory can be considered as an ability to keep informations in short-term memory and manipulate them (Borkowska 2006). During dreaming working memory, as well as self-reflective awareness and executive functions, is impoverished. If lucid dreaming is a special type of dreaming in which working memory level obtains values similar as in waking, then it may be associated with an increased activity of cortical areas attributed to working memory. According to Hobson (2001, Hobson et al. 2000) this cortical area is the dorsolateral prefrontal cortex (DLPFC).

The involvement of DLPFC in working memory has been studied in humans using neuroimaging techniques. In healthy subjects, tasks engaging working memory are associated with an increased activity of this brain region (D’Esposito and Postle 2000). Additionally, dorsolateral prefrontal cortex is one of the only cortical regions selectively deactivated during REM sleep (cf. Muzur et al. 2002). These facts led Hobson (2001) to propose the hypothesis that lucid dreaming is a result of the residual activation of DLPFC, which is amplified by the REM activation of other cortical networks. Other, subcortical structures that might be less powerfully activated during lucid dreaming are the amygdala and probably the parahippocampal cortex. In turn, the pons and the parieto-occipital junction remain activated to maintain the hallucinatory intensity of the dream (Hobson 2001).

One of the mechanisms which can facilitate the appearance of meta-awareness during dream is priming. Lucid dreaming usually occurs spontaneously, however, it can be facilitated by pre-sleep autosuggestion. Autosuggestion increases the probability of emergence of the reflective self-awareness during dreaming by priming the neural networks in prefrontal areas. A word or a sentence can prime its subsequent recall, because associated words are more easily recognized, and REM sleep can enhance this priming (Hobson et al. 2000, Hobson 2001).

The most important element of working memory is the central executive (CE), engaged every time when the stored information must be manipulated. CE is localized in the prefrontal cortex. However, there are other significant components of working memory which are distributed at different sites of the cortex. One of the working memory components, not related only to prefrontal areas, is the phonological loop, which is involved in storage and rehearsal of phonological representations. The phonological loop is localized in the parietal cortex, Broca area, left pre-motor cortex, left supplementary motor area, and right cerebellum hemisphere (Henson 2001). It seems thus probable that those regions might also be activated during lucid dreaming, and, as shown recently by Holzinger et al. (2006) and by LaBerge (1993), they are indeed activated, at least the parietal areas, Moreover, the inferior parietal lobule, considered as an important brain area involved in consciousness (Taylor 2001), is, similarly as the dorsolateral prefrontal cortex, selectively deactivated during REM sleep. As reported in a positron emission tomography (PET) study using H215O, the onset of non-REM sleep is associated with remarkable and specific deactivation of prefrontal and inferior parietal cortices. During REM sleep some prefrontal regions such as lateral orbital, dorsolateral prefrontal and opercular cortices, and also the heteromodal association cortices of the inferior parietal lobe remained attenuated (Braun et al. 1997).

It is thus more probable that the activation of parietal areas, perhaps together with dorsolateral prefrontal cortex, is necessary for lucid dreaming to occur. Baars et al. (2003) suggested that the cooperation of prefrontal and parietal cortices may have special relationship with consciousness. Crick and Koch (2003) also noted that the front part of the brain which performs the highest order processing of neural integration is interacting with the back, sensory part, resulting in conscious experience. It has been shown that although unconscious stimuli, such as visual words, activate known word-processing regions of visual cortex, the same conscious stimuli involve frontward spread of activation beyond the sensory areas (Dehaene et al. 2001, 2003). The same results have been obtained using vision, touch and pain perception and conscious versus automatic skills (Baars 2002, Baars et al. 2003). Baars et al. (2003) suggested also that “self” systems supported by fronto-parietal regions could be disabled during consciousness (including sleep).

In light of all these statements and findings, it is reasonable to suggest that increased fronto-parietal activity during sleep may lead to lucid dreaming.

Is lucid dreaming a dissociated state?

The transition among three states of vigilance (wakefulness, NREM sleep and REM sleep) is gradual, and each state consists of a number of physiologic variables. However, these primary states of being are not mutually exclusive. The physiological event markers of one state can intrude into other state or oscillate rapidly, resulting in the appearance of unconventional, bizarre behaviors, which can occur in diverse naturalistic and clinical settings (Mahowald and Schenck 2001, Mahowald 2009). Up-to-date there is a number of well-documented, the so called “state dissociations” in humans, e.g. narcolepsy, REM sleep behavior disorder (RBD), sleep paralysis, sleep inertia, out-of-body experience, hypnagogic/hypnopompic hallucinations or sleep-walking (Mahowald 2009), some of which occur spontaneously, and some appear a result of neurological dysfunctions or medication administration. Each such state dissociation is considered as a variant of the predominant or prevailing parent state (Mahowald and Schenck 2001).

Lucid dreaming is considered as a dissociated state (Hobson et al. 2000, Mahowald and Schenck 2001) in which the parent state, REM sleep, is mixed with wakefulness (Mahowald and Schenck 2001). Hobson et al. (2000) proposed that lucid dreaming is a “hybrid state lying across the wake-REM interface” and can be characterized lucid dreaming using the AIM model (Hobson et al. 2000) of brain-mind state.

The AIM model describes the three main states of consciousness observed in humans, wakefulness, NREM sleep and REM sleep, with different values of three parameters: cortical activation (“A” value), which is correlated with mind’s ability to retrieve and manipulate stored information, input source (“I” value), which shifts from external sensory input during wakefulness to internally generated signals (e.g. PGO waves during REM sleep), and neuromodulation (“M” value), defined as the ratio of aminergic to cholinergic neuromodulation. Different values of these parameters, which can be considered as the axes of the three-dimensional state space, define and distinguish conscious states (Fig. 1A). As any point within the state space can be occupied, and the parameters, A, I and M can be dissociated, this model can also describe the diversity of the dissociated states, in which some parameters obtain values characteristic for NREM sleep while other obtain values attributed to waking or REM sleep states (Hobson et al. 2000). In lucid dreaming, the dissociation is represented in the AIM model as a split AIM cube (Fig. 1B), so the part of it representing the dorsolateral prefrontal cortex is dissociated from the rest of the brain, along the “A” axis. 


Recent research by Voss et al. (2009) provided new experimental evidence supporting the hypothesis about hybrid-nature of lucid dreaming. The researchers showed showed that lucid dreaming is not only a part of REM sleep but also has some features of waking state. During lucid dreaming, as well as in REM sleep, increased power of delta and theta frequency bands were observed, accompanied with lower alpha power. Because elevated alpha power is characteristic of of the state of waking with eyes closed, this reversed pattern is an evidence that lucid dreaming occurs in a state of sleep. However, the authors also found  wake-like activation higher frequency band (~40Hz, gamma) during lucid dreaming compared to normal REM sleep, which occurred frontolaterally and frontally. In fact, increased gamma activity in lucid dreaming was already observed by Mota-Rolim et al (2008) in left prefrontal lobe, together with significantly more alpha activity in the right posterior temporal lobe, compared to non-lucid REM sleep.

The AIM model and its application to lucid dreaming has been criticized by LaBerge (2000), mostly due to its over-simplicity. As argued by LaBerge, each of three dimensions of AIM model is actually multidimensional, and in fact there is no evidence to support the idea that lucid dreaming is a dissociated state.

It may also be remarked that the statement that lucid dreaming is a mixed or hybrid state between REM sleep and wakefulness is also incorrect, or at least insufficient to describe the complete nature of this phenomenon, due to the fact that lucid dreams occur both in REM sleep and in NREM sleep. Appearance of NREM lucid dreams is observed in early sleep stages, mostly in the stage 2 (Dane 1986). According to AIM model, during the stage 2 of NREM sleep the “A” value is lower than in REM sleep. Thus, unlike as in REM sleep, residual activation of DLPFC cannot be amplified by the REM activation from other cortical areas. In fact, if the decrease of DLPFC activity progresses with the deepening of NREM sleep, and remains deactivated in REM sleep (cf. Muzur et al. 2002), then residual activation of DLPFC should be greater in the stage 2 of NREM sleep than during REM sleep (in other words, DLPFC is not fully deactivated). In this case, another possible mechanism, which could lead to re-activation of DLPFC and, consequently, to lucid dreaming during early NREM sleep, should be proposed.

Neuroimaging of Lucid Dreaming

...coming soon




References (in alphabetical order):

1. Aristotle (350 B.C.) “On Dreams”, Translated by J. I. Beare, access online (11.05.2009): http://classics.mit.edu/Aristotle/dreams.html.

2. Aserinsky E, Kleitman N (1953) Regularly occurring periods of eye motility, and concomitant phenomena during sleep. Science 118: 273-274.

3. Baars BJ (2002) The conscious access hypothesis: origins and recent evidence. Trends Cogn Sci 6: 47-52.

4. Baars BJ, Ramsøy TZ, Laureys S (2003) Brain, conscious experience and the observing self. Trends Neurosci 26: 671-675.

5. Borkowska A (2006) Neuropsychological and neurobiological aspects of working memory (in Polish).Neuropsychiatria i Neuropsychologia 1: 31-42

6. Braun AR, Balkin TJ, Wesensten NJ, Carson RE, Varga M, Baldwin P, Selbie S, Belenky G, Herscovitch P (1997) Regional cerebral blood flow throughout the sleep-wake cycle. An H215O PET study. Brain 120: 1173-1197.

7. Callaway CW, Lydic R, Baghdoyan HA, Hobson JA (1987) Pontogeniculooccipital waves: spontaneous visual system activity during rapid eye movement sleep. Cell Mol Neurobiol. 7:105–49.

8. Cantero JL, Atienza M, Salas RM, Gómez C (1999) Alpha power modulation during periods with rapid oculomotor activity in human REM sleep. NeuroReport 10: 1817-1820.

9. Cigogna P, Bosinelli M (1994) Models of consciousness and mental experience during sleep. J Sleep Res 3(Suppl 1): 44.

10. Cigogna P, Bosinelli M (2001) Consciousness during dreams. Conscious Cogn 10: 26-41.

11. Crick F, Koch C (2003) A framework for consciousness. Natl Neurosci 6: 119-126.

12. D’Esposito M, Postle BR (2000) Prefrontal cortical contributions to working memory: evidence from event related fMRI studies. Exp Brain Res 133: 3-11.

13. Dane J (1986) Non-REM lucid dreaming. Lucidity Lett 5(1): 133-147.

14. Dehaene S et al (2001) Cerebral mechanisms of word masking and unconscious repetition priming. Natl Neurosci 4: 752-758.

15. Dehaene S et al (2003) A neuronal network model linking subjective reports and objective physiological data during conscious perception. Proc Natl Acad Sci USA 100: 8520-8525.

16. Dietrich A (2003) Functional neuroanatomy of altered states of consciousness: the transient hypofrontality hypothesis. Conscious Cogn 21: 231-256.

17. Erlacher D, Schredl M (2008) Do REM (lucid) dreamed and executed actions share the same neural substrate? Int J Dream Res 1(1): 7-14.

18. Erlacher D, Schredl M, LaBerge S (2003) Motor area activation during dreamed hand clenching: a pilot study on EEG alpha band. Sleep and Hypnosis 5: 182-187.

19. Fosse MJ, Fosse R, Hobson JA, Stickgold RJ (2003) Dreaming and episodic memory: a functional dissociation? J Cogn Neurosci 15: 1-9.

20. Green C, McCreery C (1994) Lucid Dreaming. London: Routledge.

21. Henson R (2001) Neural working memory. In: Working memory in perspective. Andrade J, Psychology Press, 151-173.

22. Hervey de Saint-Denys (1867) „Dreams and How to Guide Them”. Translated by Nicolas Fry, edited and with an introduction by Morton Schatzman. London: Duckworth, 1982.

23. Hirsch TB, Creem SH, Wraga M, Harrington GS, Fox KV, Profitt DR, Downs JH III (1999) Imagined rotations of the self: an fMRI study. NeuroImage 9: S803.

24. Hobson JA (2001) The dream drugstore. Cambridge, MA. MIT Press, 96-98.

25. Hobson JA, McCarley RW (1977) The brain as a dream state generator: an activation-synthesis hypothesis of the dream process. Am J Psychiatry 134:1335–48.

26. Hobson JA, Pace-Schott EF (2002) The cognitive neuroscience of sleep: neuronal systems, consciousness, and learning. Nat Rev Neurosci. 3:679–93.

27. Hobson JA, Pace-Schott EF, Stickgold R (2000) Dreaming and the brain: toward a cognitive neuroscience of conscious states. Behav Brain Sci 23: 793-842.

28. Holzinger B, LaBerge S, Levitan L (2006) Psychophysiological correlates of lucid dreaming. Dreaming 16: 88-95.

29. Iacoboni M (2009) Imitation, empathy and mirror neurons. Annu Rev Psychol. 60: 653-670.

30. Kahan TL (1994) Measuring dream self-reflectiveness: a comparison of two approaches. Dreaming 4(3): 177-193.

31. Kahan TL, LaBerge S (1994) Lucid dreaming as metacognition: implication for cognitive science. Conscious Cogn 3: 246-264.

32. Kahan TL, LaBerge S, Levitan L, Zimbardo P (1997) Similarities and differences netween dreaming and waking cognition: an exploratory study. Conscious Cogn 6: 132-147.

33. Kales A, Hoedemaker FS, Jacobson A, Kales JD, Paulson MJ, Wilson TE (1967) Mentation during sleep: REM and NREM recall reports. Percept Mot Skills 24:555-60.

34. Kaufman L, Schwartz B, Salustri C, Williamson SJ (1990) Modulation of spontaneous brain activity during mental imaginery. J Cogn Neurosci 2: 124-132.

35. LaBerge S (1990) Lucid dreaming: psychophysiological studies of consciousness during REM sleep. In: Bootzen RR, Kihlstrom JF, Schacter DL (Eds.) Sleep and Cognition. Washington D.C.: American Psychological Association, pp.: 109-126.

36. LaBerge S (1993) Lucidity research, past and future. NightLight 5: 329-335.

37. LaBerge, S. (2000). Lucid dreaming: Evidence and methodology. Behav Brain Sci 23(6): 962-963.

38. LaBerge S, Dement WC (1982) Lateralization of alpha activity for dreamed singing and counting during REM sleep. Psychophysiology 19: 331-332.

39. LaBerge S, Levitan L, Dement WC (1986) Lucid dreaming: physiological correlates of consciousness during REM sleep. J Mind Behav 7: 251-258.

40. LaBerge S, Nagel L, Dement WC, Zarcone V (1981) Lucid dreaming verified by volitional communication during REM sleep. Percept Mot Skills 52: 727-732.

41. LaBerge S, Rheingold H (1990) Exploring the world of lucid dreams. New York, Ballantine.

42. Lim AS, Lozano AM, Moro E, Hamani C, Hutchison WD, Dostrovsky JO, Lang AE, Wennberg RA, Murray BJ (2007) Characterization of REM-sleep associated ponto – geniculo – occipital waves in the human pons. Sleep 30(7): 823-827.

43. Mahowald MW (2009) What state dissociation can teach us about consciousness and the function of sleep. Sleep Med 10: 159-160.

44. Mahowald MW, Schenck CH (2001) Evolving concepts of human state dissociation. Arch Ital Biol 139: 269-300.

45. Maquet P, Peters J, Aerts G, Delfiore C, Degueldre A, Luxen A, Franck G (1996) Functional neuroanatomy of human rapid-eye-movement sleep and dreaming. Nature 282 (6596): 163-166.

46. Moruzzi G (1963) Active processes in the brain stem during sleep. Harvey Lect. 58: 233-297.

47. Muzur A (2005) Toward an integrative theory of sleep and dreaming. J Theoretical Biology 233: 103-118.

48. Muzur A, Pace-Schott EF, Hobson JA (2002) The prefrontal cortex in sleep. Trends Cogn Sci 6(11): 475-481.

49. Nielsen TA (2000) A review of mentation during REM and NREM sleep: “covert” REM sleep as a possible reconciliation of two opposing models. Behav Brain Sci 23: 851-866.

50. Nofzinger EA, Mintun MA, Wiseman M, Kupfer DJ, More RY (1997) Forebrain activation in REM sleep: an FDG PET study. Brain Res 770(1-2):192-201.

51. Ogilvie RD, Hunt HT, Tyson PD, Lucescu ML, Jeakins DB (1982) Lucid dreaming and alpha activity: a preliminary report. Percept Mot Skills 55: 795-808.

52. Ogilvie RD, Vieira KP, Small RJ (1988) EEG activity during signaled lucid dreams. ASD Newsl 5: 5-6.

53. Piller, R. (2009). Cerebral specialization during lucid dreaming: A right hemisphere hypothesis. Dreaming, 19 (4), 273-286.

54. Posner MI, Rothbart MK (1998). Attention, self-regulation and consciousness. Phil Trans R Soc Lond B 353: 1915-1527.

55. Ramachandran V (2007) The neurology of self-awareness. Edge 201, access online (25.05.2009): www.edge.org.

56. Rechtschaffen A (1978) The single-mindedness and isolation of dreams. Sleep 1: 97-109.

57. Rizzolatti G, Craighero L (2004) The mirror-neuron system. Annu Rev Neurosci 27: 169-192.

58. Schupp HT, Lutzenberger W, Birbaumer N, Miltner W, Braun C (1994) Neurophysiological differences between perception and imaginery. Cogn Brain Res 2: 77-86.

59. Simons JS, Spiers HJ (2003) Prefrontal and medial temporal lobe interactions in long-term memory. Nat Rev Neurosci. 4: 637-648.

60. Strelen J (2006) Acoustic evoked potentials in lucid dreaming (in German). Ph. D. thesis, University of Mainz.

61. Suzuki H, Uchiyama M, Tagaya H, Ozaki A, Kuriyama K, Aritake S, Shibui K, Tan X, Kamei Y, Kuga R (2004) Dreaming during non-rapid eye movement sleep in the absent of prior rapid eye movement sleep. Sleep 27: 1486-1490.

62. Taylor JG (2001) The central role of the parietal lobes in consciousness. Conscious Cogn 10: 379-417.

63. Tyson PD, Ogilvie RD, Hunt HT (1984) Lucid, prelucid and non-lucid dreams related to the amount of EEG alpha activity during REM sleep. Psychophysiology 21: 442-451.

64. van Eeden F (1913) A study of dreams. Proc Soc Psychical Res 26: 431-461.

Disclaimer

The images contained within this site were sourced via Flickr.com and Google Images and are part of a Creative Commons-licensed content . Each image has a link to its source. 

If, for any reason, the owner of an image wishes to have their image removed from this site, please send me an email informing about it.


All articles published at lucidologia.blogspot.com are written by Andrzej Wnuk (email).
The Science of Lucid Dreaming was first published (24 November 2010) at dreaminglucid.blogspot.com.
Last update of The Science of Lucid Dreaming: 20 April 2011.
Copyright (C) 2010-1011 Lucidologia