Chapter
Fifteen
Dreaming and beyond
are not points, however, but more like clouds in
state space, especially the waking state in which
the values are all high but change from moment
to moment.
Things may not, however, be this simple. First, there
is the obvious point that the map is crude, includ-
ing only three dimensions, while the reality is much
more complicated (LaBerge, 2000; Solms, 2000).
This is not a serious problem since the scheme still
provides a way of relating sleep states to other sup-
posed ASCs and to the various neurotransmitter
systems that control the overall state of the brain,
and more detail and further dimensions could
potentially be added.
More troublesome is that the correlation between
REM and dreaming, while real enough, is not per-
fect. In the early days of sleep research, REM sleep
and dreaming were often treated as equivalent, but
subsequently people became more careful in refer-
ring either to the physical state or to the reported
experiences. Dreaming is reported in about 70–95%
of awakenings from REM sleep and roughly 5–10%
of non-REM sleep, while mentation of some sort is
reported from about 50% of non-REM awakenings.
For example, after repetitive activities like playing
a skiing game for two hours over a few days, many
people woken shortly after falling asleep reported
images clearly related to the game (Wamsley et al.,
2010). In the skiing case, participants also reported
images from past skiing experiences (especially
crashes); if woken later in sleep, the imagery reported
was more remote from the game, like stacking wood
at a ski resort. The finding that non-REM experiences
may become more ‘dreamlike’ as sleep continues
raises the question of where exactly the boundary
between ‘dreaming’ and ‘sleep mentation’ should be
set: should we allow mentation to be any mental
activity (e.g. perceptions, bodily feelings, thoughts)
but try to restrict dreaming to ‘more elaborate, vivid,
and story-like experiences recalled upon awakening’
(Kryger, Roth, and Dement, 2011, p. 585), or are such
distinctions arbitrary and impossible to apply consis-
tently? Maybe even REM and non-REM themselves
cannot be neatly distinguished, and non-REM sleep
might include covert REM processes (Nielsen, 2000).
Overall, though, it is clear that being physiologically
in REM sleep does not guarantee dreaming, and
dreaming can occur without the physiological state
of REM.
mechanisms, and ASCs. He has posts at both the Uni-
versity of Turku in Finland and the University of Skövde,
Sweden. As a Harry Potter fan, he claims to use Professor
Dumbledore’s Elderwand as a pointer, but we cannot
confirm rumours that he was once a visiting professor at
Hogwarts, specialising in Defence Against the Dark Arts.
Revonsuo is best known for his evolutionary theory of
dreaming as threat simulation, and his advocacy of the
dreaming brain as a model for understanding conscious-
ness. He describes himself as a ‘biological realist’, and
believes that consciousness is a higher level of biological
organisation in the brain. In future, he plans to work
more on visual consciousness and anaesthesia.C
on
C
e
P
t
15.1
tHe eVoLUtIon oF DReAmInG
Why did dreaming evolve? the question
here is not why sleep evolved. there are
many competing theories about the evolu-
tionary functions of Rem and non-Rem sleep
in different species (Horne, 2006; Barrett
and mcnamara, 2012; Hobson and Friston,
2012), but the trickier question concerns
dreams: do they have a function of their
own or are they an inevitable concomitant
of certain sleep states? As with the question
of the evolution of consciousness itself (Chap-
ter 11), we can find examples of all the main
approaches.
Dreaming may have a crucial biological func-
tion. According to Antti Revonsuo’s threat sim-
ulation theory, during most of human evolu-
tion serious physical and interpersonal threats
meant a reproductive advantage for those who
survived them, so dreaming evolved to simu-
late and practise dealing with these threats (an