Biology 202
1999 Second Web Reports

Proposed Mechanisms of Dreaming

Feyza Sancar

New physiological discoveries made in the 1950's linked a particular phase of sleep with dreaming (8). This phase of sleep is known as the REM (rapid eye movement) phase. This newly acquired information spawned refreshed interest in the mechanisms (specifically neurophysiological mechanisms) of dreaming. Validity of the physiological and neurobiological approach to dreaming was supported by certain (current) clinically measured and observed behaviors accompanying REM sleep (8). These behaviors or characteristics include:

-phasic clusters of extraocular muscles of the eye producing rapid eye movement
-generalized activation of the forebrain (cerebral cortex)
-phasic activation of the visual pathway
-inhibition of sensory input
-suppressed motor activity
-activation or inhibition of various brain stem neurons. (5)

Many of these behavioral markers associated with REM sleep closely dictate or reflect the dream phenomenon. Although the REM sleep phase and the dream state are closely related, REM sleep is not necessary for dreaming. However the prevalence of dreams are certainly greater in REM sleep (3). Some researchers also contend that REM dreams are uniquely different from those reported in non-REM dreams both in content and quality (11,12). Reports from REM sleep awakenings are typically longer, more vivid, and more emotionally charged than non REM sleep reports (2). Non REM sleep reports also reflect a more thought-like rumination, concerned with realistic, "common place" events (2,4,12). For many researchers, REM dreaming is considered the most elaborate or 'true' form of dreaming, especially with regards to the intense level of activation in the brain (2,4,5).

According to Alan Hobson and Robert McCarley of Harvard University, there is a clearly plausible neurophysiological explanation for the dreaming process. In their "Activation-Synthesis" model, they implicate the brain stem as the location of the "dream state generator". According to this theory, the brain stem produces REM periods, as well as triggering the dream state. During these REM dream states, sensory input and motor output is blocked, while the cerebral cortex is activated by the partially random impulses received from the brain stem (1,2,8). Presumably, the cerebral cortex (forebrain), once activated, proceeds to process and synthesize the internally generated information received from the brain stem (1,2). It is this series of activation and synthesis which creates the imagery we experience in the dream state. The forebrain receives the random signals from the brain stem and begins to assimilate the information into a coherent pattern. Presumably, the information received by the forebrain is spatially specific (hence the occurrence of 'rapid eye movement') and genetically programmed (1).

This reference to brain stem impulses as genetically programmed alludes to the concept of central pattern generators. It is known that central pattern generators exist for walking. The particular preprogrammed neural information produced by the brain stem during dream states may also be a product of a type of pattern generator. Instead of producing observable behaviors, such as walking, they may ultimately create internally generated (an thus not outwardly observable) behaviors such as visual and auditory hallucinations. It does, however, seem as though the link between the behavior and the pattern generator is not as direct since the forebrain plays an important role in both creating and synthesizing images. The fundamental pieces of information used in producing the final image may, however, be dependent on several central pattern generators. This is especially plausible since the behavioral outcome observed during the dreaming state (hallucinations) are present in the absence of sensory input.

To reiterating, it is the forebrain, however, that must then successfully integrate or couple this random, genetically programmed information with the appropriate experiential data contained with in its memory banks. Consequently, the forebrain is not always capable of conjuring memories which complement the relatively incomplete primary stimuli received from the brain stem. The frequent 'mismatching' of primary inputs of the brain stem and experiential data provided by the forebrain results in the commonly experienced bizarre quality of dream imagery. This phenomenon accounts for the presence of dream characters which seem to be an amalgam of objects, people, and animals previously encountered in the 'real' world-i.e.-a dog with your grandmothers head and a string of sausage links for a tail. This phenomenon is also thought to explain other such "bizarre formal qualities of dream mentation" including discontinuous scene shifts, and symbol formation. (1,2)

As the activation synthesis model provided a very viable neurophysiological approach to dreaming, the model was quite abstract and presented many unanswered questions. Firstly, why is the brain stem randomly activated during the dream state? Secondly, what exactly activates or disinhibits the brain stem? Further more, what are the specific structures and pathways involved in the process of dreaming? As the study of sleep and dreaming progressed, a very detailed network of neuronal activation patterns was determined. As the Activation-Synthesis model suggested, structures in the brain stem are intimately involved in the dream state (1). Interestingly, the triggering mechanism responsible for the neuronal genesis of dreaming was soon linked to the reciprocal interaction of neuromodulatory systems in the brain stem (3). These neuromodulatory systems are comprised of noradrenegergic and serotonergic neurons of the Locus Coeruleus (LC--found in the dorsal pons) and nucleus of Raphe (found in the reticular formation), as well as the Acetylcholinergic neurons (found in the peribrachial area of the reticular activating system) (2,13). In terms of dreaming and REM sleep, aminergic systems are inactive during this period and cholenergic systems are active. The characteristic inability to remember many dreams is a direct effect of the dynamics of the neuromodulatory systems (2).

Currently, it is surmised that the activity of noradrenergic and serotonergic neurons of the LC and Raphe nucleus are suppressed upon entrance into REM states (2,13). Their suppression leads to the cessation of both aminergic modulation and inhibition in the brain stem oculomotor networks (allowing for eye movement), geniculate bodies (in the thalamus), and the visual cortex (2). Initially, the disinhibition of these structures located in the brain stem causes peribrachial cholinergic neurons of the pons (also in the brain stem) to become hyperexcitable, and hence fire in synchronous bursts. These neuronal bursts of the pons Cholinergic neurons propagate through (and phasically activate) the geniculate bodies in the thalamus, and the visual cortex. The propagation through these three bodies can be experimentally measured and are referred to as PGO (pontogeniculoocciptal) waves. (2,13)

These waves constitute the internal stimulus source for the visual system in REM sleep, as well as simultaneously providing information about the direction of eye movements. The cholinergic neurons also excite neurons in the medial pontine reticular formation (mPRF), that in turn activate neurons of the forebrain (13). It is the activation of the forebrain that is responsible for the cortical activity that accompanies REM sleep. Presumably, this cortical activation also constitutes the ability to form a relatively coherent dream plot from the inputs regarding visual and directional information (2,13). Furthermore, the general shift from aminergic to cholinergic systems has been reported to increase the noise level of cortical neurotransmission, as well as the ease of attentional shifts in perceptual cueing tasks. Subsequently, this may account for the bizarre features of dreaming (2) since there would be many more random inputs to make 'sense' of. Interestingly, activation of the geniculate bodies of the thalamus, as well as their interaction with the cerebral cortex, may also play an important role in the visual components as well as auditory components of dream plots (4).

Firstly, the activated cortex during REM dreaming drives the thalamus to generate activity that contributes to dream sensations (4). The cortex also prevents the thalamocortical system from generating representations of events at the sensory periphery (4). As such, it is surmised that complex imagistic representations in dream states require adequate functioning of relevant cortical areas, as well as sensory-specific thalamic areas (4). Clinical observations support this claim. For example, cortical lesions in prosopagnosics (individuals incapable of recognizing faces) also lose the capacity to dream faces (4). Similarly, selective lesions of the lateral geniculate nucleus of the thalamus leave patients with visual scotomas (dark or blind spots) in their wakeful visual perceptions as well as in dream imagery (4). This seems to imply that the thalamus is involved in producing or constructing dream imagery. Interestingly, schizophrenics show increased thalamic activity during auditory hallucinations (4). Perhaps this implies that with extreme activity of the thalamic area (particularly in relation to 'normal' waking subjects) sensory perceptions can in fact be internally generated. Although this increased neuronal activity of the thalamic area is expected or 'normal' in REM-dreaming states, it is pathological in any kind of waking state.

On a similar vein, it has been observed that a small population of schizophrenics displayed an increase in the number of neurons in an area of the brain stem which is part of the Reticular Activating System (RAS) (9). As it has been mentioned, certain areas of the RAS contain cholinergic neurons responsible for the activation of the cerebral cortex (particularly the visual cortex) during the REM dream state. Perhaps this observation implies that an increase in the number of Acetylcholine releasing neurons (hence an increase in the amount of acetylcholine being released) is capable of causing an abnormal or inappropriate hyperarousal of certain areas in the cerebral cortex during waking states. In other words, cholinergic systems, when inadequately modulated (or unbalanced) during the waking state can cause similar hallucinatory episodes experienced during the REM dream state. These observations and possible explanations further implicate areas of the RAS in the internal, sensory independent, production of images characteristic of the dream state. It is noteworthy here to mention that, just as in the associations between dreaming and schizophrenia, dreams have been frequently viewed as a model for waking forms of mental illness and psychosis (12).

Although the discussed pathways and structures implicated in dreaming provide relatively clear processes involved in imagery, the mechanisms behind emotional content of dreams are not as clear. It has been concluded that the limbic systems largely related to emotional states are specifically connected to the structures active during dreaming sleep (3). Therefore, through the law of physical contiguity, once the brain stem is activated, it propagates signals to successive neural pathways that are ultimately connected to parts of the brain responsible for emotions. As a supporting example, physiological observations have reported increased blood flow to the amygdala during REM-dreaming (5).

This observation has been seen, by many researchers, an indication of the basis for the emotional content in dreams (5). Others extrapolate on this notion to propose that perhaps the activation of these structure involved in emotion is not a cause of the emotional content in dreams, but a product of the necessity of these emotions (5). It has been suggested that 'feelings' in dreams are caused by a "narrative necessity". The feelings are created with the goal of being generally appropriate to the immediate narrative context of the unfolding dream scenario (5). This again implies the role of the cerebral cortex in organizing and constructing a sufficiently coherent 'story' to compliment the inputs received by parts of the brain stem. The emotional content may be mediated by the cortex which determines the emotional response best suited for the context of the imagery and 'dream scenario' being created. Perhaps this mechanism is a product of a type of 'cognitive dissonance' where the presence of a plethora of images does not seem complete without their complementing emotions. As a result, the cerebral cortex uses its 'executive' functions to supply the 'missing' information.

The presented information has outlined a predominantly neurobiological/neurocognitive approach to the mechanisms behind dreaming. There are still many alternate views on dreaming (5,6,12). There are especially contrasting arguments on the subsequent function of dreams. The neurobiological approach recognizes the importance of REM sleep, but contends that people do not particularly 'need' dreams (6). They are mainly viewed as a way to exercise neuronal circuitry to ensure efficiency and efficacy (2). As such, the process involved in producing dreams may be important, but the dreams themselves may not (2,6). As it can be seen, there is still room for investigation in the field of dream neurobiology and physiology. There are in fact currently impending experiments reflecting new or continuing concerns (10). There are many questions that the neurobiological concepts fails to address, such as the possibility of conscious decision making during the dream state (i.e.-lucid dreaming) (1,2). The possibility of the I-function being actively involved in the dreaming process provides intriguing possibilities. With these continually arising questions and implications, as well as the established neurobiological correlates, a new and perhaps never ending chapter in dream research can be established.

WWW Sources

1)Dreaming: Function and Meaning, by Dr. LaBerge

2)Dreaming:A Neurocognitive Approach, by Allan Hobson and Robert Stickgold

3)A Study of the Neurophysiological Mechanisms of Dreaming, by M. Jouvet and D. Jouvet

4)Dreams and Other Imagery

5)A Contemporary Neurobiology of Dreaming?, by David Foulkes

6)A physiology of Dreams?: Complete article


8)A Physiology of Dreams?: REM side bar

9)Neurobiology Research

10)2166 NIA-Basic Clinical Research on Sleep and Wakefulness

11)Brain/Body Activity During Sleep and Dreams

12)Paradigms of Consciousness During Sleep

13) Carlson,Neil (1998). Physiology of Behavior: Sixth Edition. Allyn and Bacon. Needham Heights, MA.