[ CogSci Summaries home | UP | email ]


Hobson, J. A. & McCarley, R. W. (1977). The brain as a dream state generator: An Activation- Synthesis Hypothesis of the dream process. The American Journal of Psychiatry, 134:12, 1335-1348.


    author =           {J. Allen Hobson and Robert W. McCarley},

    title =                {The brain as a dream state generator: An Activation- Synthesis Hypothesis of the dream process},  

    journal =          {The American Journal of Psychiatry},

    year =               {1977},

    volume =         {134:12},

    pages =            {1335—1348}



Author of the summary: Natalie Poirier, 2012, npoirie1@connect.carleton.ca


Cite this paper for:

-          Activation-Synthesis Hypothesis: neuro-physiological evidence for the process and production of dreams [p2]

-          Dreams occur during desynchronized sleep, are visual and vivid, and have state-dependent amnesia

-          Pontine brain stem is critical region for the creation of dreams

-          There is sensory input and motor output, but motor output is blocked by spinal cord neurons

-          Giant cells within the pontine reticular formation are key within almost processes involved during sleep. [p8]


Dream theory, once known through Freud’s psychoanalytical theory of the mind, is now showing to be more physical than once believed. A new theory being looked at is called the Activation-Synthesis Hypothesis. This hypothesis looks at the regularly reoccurring physical brain state that is called ‘dreaming sleep’. Also looks at how the brain is able to simultaneously block and take in inputs and outputs (mainly sensory input and motor output). They had hypothesised that the “generated sensorimotor information, which is partially random and partially specific, is then compared with stored sensorimotor data in synthesis of dream content.” [p2]


What is a dream?

Described as a mental experience, a dream happens during sleep and is characterized by hallucinoid information. They are generally accepted as real and are mostly visual and vivid. [2] Dreams usually feel like they are real and can potentially create strong emotions. They are also quite difficult to recall unless you wake up directly from REM sleep.


What is the state of the brain during dreaming sleep?

There are three states that the Central Nervous System (CNS) can be in: waking (W), synchronized sleep (S) and desynchronized sleep (D). The way to differentiate between these states is to look at EEGs, EMGs, and electrooculograms. What they found was that dreams most frequently take place during REM sleep. What makes D sleep different from the other states is the activation of the brain, which generally excludes external input but still has the generation of internal input coupled with the blocking of motor output.


Electrophysiology of the brain during the dream state

There are three major physiological features of D sleep that have to be looked at during this state to understand what is happening within the brain. [p3]

1.       How is the forebrain activated in the D state?

a.       After looking as EEG scans of the brain, it is evident that the reticular formation of the anterior brain stem is as activated during the waking phase as it is during the dreaming one.

2.       How is motor output blocked in the D state?

a.       It is shown that motor inhibition is not hindered within the brain itself but stopped from being expressed by the spinal cord motor neurons

3.       How is sensory imagery generated in the D state?

a.       There is a connection with the waking state in that there also seems to be a corollary discharge of the occulomotor system, which stops visual broadcast during saccadic eye movement. There is the possibility that the intensity of the eye movements during dreaming dictates how intense the dream itself is

From these, we can see that there is a connection between D state and W state that allows to conclude that D state is the one in which dreaming occurs because there is still a lot of brain activation that does not appear in S sleep.

The dream state is found to occur periodically throughout the night and seems to be a pre-programmed event.


When animals were studied, it was found that they had different length of sleep cycles depending on their species.


Localization of the power supply or trigger zone of the dream state generator

It is found that the pontine brain stem is critical to creating dreams. [p4] When there was damage to the pontine reticular formation, D sleep was prevented from occurring. This would lead to the connection that it may be the triggering mechanism for dreaming. There is also a connection between body temperature and the periodicity of dreams. A lesion within the pontine brainstem also removes atonia during D sleep.

What these findings imply is that:

1.       The forebrain is activated during dreaming

2.       The spinal reflexes are inhibited, which would therefore prevent output of motor activities regardless of the level of activity within the brain

3.       The eyes and the resticular systems are activated by pontine reticular formation; and

4.       Internal information is generated by activation of various motor systems [p6]


Histological features of relevance to the periodic triggering of the dreaming sleep state generator

Taken from Cojal’s discussion of histology, he emphasizes three main points: [p6-7]

1.       The paramedian reticular giant cells influence is found to cause excitation in other cells, although diminished from the level in which it is seen in paramedian reticular giant cells.

2.       Raphe neurons are situated and connected to manage excitability of paramedian elements

3.       A small lateral zone of stellate cells may be the transmitter of inputs to the central reticular core. This may be an important feature that shows a dependency with adaption on the ability to interrupt a cycle and not to allow all the exogenous stimuli into the dream plot.


Cellular activity in the pontine brain stem during the sleep cycle

There are three different ways that the importance of the pontine brain stem at the level of the single cell is affected during D sleep. [p8]

1.       Selectivity Criterion: which cells change rate most in D?

a.       The cells that played a controlling role with other cells were the ones that had an extremely altered discharge rate over the sleep cycle. These cells seemed to be the giant cells of the pontine tegmentum that had a greater amount of discharge during D sleep potentially initiating dream sleep.

2.       Tonic latency Criterion: which cells change rate first in D onset?

a.       Giant cells were also shown to have a significant increase in firing rate for as long as five minutes before the onset of D sleep. This was also the maximum excitability that the giant cells could reach.

3.       Phasic latency Criterion: which cells fire before eye movements of D?

a.       To test the difference in rates, giant cells were tested in both wake and sleep states. On average, the firing rates were higher and anticipated eye movements by longer intervals than any other brain stem neurons [p9]

There are also some points that were discovered during the experimentation on the pontine brain stem.[p 8]

1.       Periodicity Criterion

a.       Giant cells recordings showed that their highest firing rate was during the full blown D episodes. The periodicity of the discharges help to show the underlying cell activity, that sleep cycles are periodic and that firing rates are controlled by the neurobiological clocks.

2.       Phasic pattern Criterion

a.       D episodes show that the pacemaker mechanism does not play a role in the excitation and the firing of the giant cells.

3.       Reciprocal interaction Criterion

a.       If giant cells are not the only cell that affects D sleep, and there is an inhibitory cell group, which cell group is it and in what way would that reciprocal regulation be effected. It is believed that it is ruled by an unknown posterior locus coeruleus cell which would do the opposite of the giant cells within the pontine brain stem (inhibit)


Model for the brain stem sleep cycle oscillator

It was found that microinjections of cholinomimetic substance carbochol into the giant cells would give a more excited D phase than injections into any other zone. [p10] It is also significant to see that an injection at locus coeruleus would create an intense inhibitory effect (as though the inhibitory cells were being activated). The distance between the cell fields also seems to have an effect on the phase lag between reciprocal cycles.


The physiological model of the dream process is also paralleled by math models which gives weight to the finds on the Activation-Synthesis Hypothesis.


Psychological implications of the cellular neurophysiology of dream sleep generation

The implications and findings of the Activation-Synthesis Hypothesis are: [p12-13]

1.       Primary motivating force for dreaming

a.       This hypothesis shows that dreams are not so much psychologically driven but the result of physiological processes.

2.       Specific stimuli for dream imagery

a.       Stimuli comes intracerebrally from the pontine brain stem, not from the cognitive areas within the cerebellum

b.      Semi-coherent imagery is produced from random signals what are sent up from the brain stem.

3.       Elaboration of brain stem stimulus

a.       The role of the perceptual, conceptual and emotional structures of the forebrain is now seen more as a synthetic constructive process and no longer as the primary distortion of a dream as Freud thought)

b.      Dreams are created synthetically and transparently, not degradative and opaquely.

4.       With respect to the forgetting of dreams

a.       Not being able to recall a dream after it seems to suggest a state-dependent amnesia.


Summary author’s notes:

-          Page numbers are from a printed handout


Back to the Cognitive Science Summaries homepage
Cognitive Science Summaries Webmaster:
JimDavies (jim@jimdavies.org)