The New Neuropsychology of Sleep Commentary by Allen Braun

Allan Hobson has presented what is essentially a modification of his activation-synthesis model; Mark Solms reviews this and presents a novel and interesting model of his own. In some sense this debate might have been far more lively five years ago, for there is now much room for common ground between these two. Nevertheless there is an air of contentiousness evident which I will argue may be simply due to the specter of Freud, who is never far from the center of this dialogue.

There are some problems with Hobson's new model: activation-input-mode (AIM) appears to treat the conditions it uses to define states of consciousness—levels of activity, sensitivity to internal vs. external outputs, and neuromodulation—as if they were orthogonal. In fact they undoubtedly interact in complex ways that make this linear model a bit too simple. Nevertheless, the model, at least as it is used to illustrate REM in Figures 3 and 4, is much more flexible than activation-synthesis and, with respect to “activation” in any case, is no longer entirely brainstem based, but acknowledges a more complex and active role for the forebrain. At the same time, in much of what he writes, Solms seeks to show that dreams can be uncoupled from REM in order to liberate the dream process from brainstem mechanisms. I suspect that this essentially represents a response to the initial activation-synthesis hypothesis—in which the dream was a secondary attempt by the forebrain to “synthesize,” i.e., create order out of its regular but chaotic activation by the pontine reticular formation. Dreams were in essence meaningless and psychologically empty.

In the present version—modified in light of Solms's own clinico-anatomical data and PET studies conducted by us, as well as by Maquet's group in Belgium (Maquet et al., 1996), Nofzinger's in Pittsburgh (Nofzinger et al., 1997), and Madsen and coworkers in Denmark —Hobson no longer dismisses dream content as vacuous. The “synthetic”