Neurobiological Theories of Dreaming

Theory 1: ACTIVATION-SYNTHESIS HYPOTHESIS

A01: There is random electrical activity during sleep: this is experiences as a dream. Dreams have no intrinsic meaning - meaningfulness is not what drives the dream.

A02: Supported by... PET scans showing that the brainstem is active during REM sleep (supporting the activity part of the model), and that the prefrontal cortex is inactive during REM sleep (supporting the synthesis part of the model) (BRAUN et al., 1997)

A01: ACTIVATION

• The brainstem generates spontaneous random signals during REM sleep.
• The boyd is paralysed so there is no other sensory input or motor output.
• For the cortex, these signals are indistinguishable from the internal and external stimuli that are usually process.

A01: Synthesis

• When activity arising from the brainstem is mixed with stored images from memory, the brain attempts to 'interpret' these, often resulting in bizarre images which we experience as dreams.

A01: Challenging evidence

• SOLMS (2000) found that in patients with damage to the brainstem, REM activity ceased but dreaming continued, suggesting that REM activity and dreaming may be independent.

• Some studies have found that many dreams are coherent and consistent over time (DOMHOFF, 1996)

A02: An alternative explanation is...

The AIM Model, (HOBSON, 1992), which includes activation of the brain, external or internal inputs and modulators. Waking, REM and NREM states vary in levels of A, I and M.

Theory 2: REVERSE LEARNING

A01: The brain requires memory consolidation to use space more efficiently. Consolidation takes place during sleep, and the associated electrical activity is experienced as a dream.

A02: Research Support

• GOERTZL (1997) - even allowing for compression of information, the brain takes in far more than it can store.
• Mammals without REM sleep have a disproportionately large cortex to compensate for the extra storage capacity needed. (MUKHAMETOV, 1987)

A01: Reverse learning

• This makes important memories more accessible, and makes storage in neural networks more compact, without the memory overlap caused by unwanted associations.
• This is adaptive because we can have smaller, more efficient brains.

A01: Neurobiological mechanism

Random stimulation of the forebrain by the brainstem excites neurons corresponding to the unwanted associations. Their synapses are modified to make future activation less likely.

A02: Challenging evidence

• Brains of mammals that lack sleep have cortexes that are not as highly folded as human cortexes so may not have a much larger capacity.

• CHRISTOS (1996) found that computer simulations of reverse learning have shown that the number of parasitic memories is increased not decreased.

Main Problem: No One neurological activity is equivalent to the state of dreaming.