[Martin Taylor 920929 16:50]
(Rick Marken 920928.1030)
I guess no other fool has rushed into the psychodynamic gritty slimy stuff
dumped into CSG-L by our resident loose canon, so maybe I'll try a first slip.
These people are particularly interested in dreams,
literature and art as sources of information about our essential selves.
One of my earlier reasons for getting interested in PCT is that it seemed
to give a very straightforward explanation of both dream structure and
dream content. Since I discovered the degrees-of-freedom argument for
the coexistence of massively parallel passive processing and active sequential
control processing, the relevance of PCT seems even more pronounced. Dreams,
in this view, are not the same as planning processes, but they are closely
related. One would expect any organism to dream that based its interaction
with the world on a wide-ranging sensory system and a low-bandwidth effector
system, and that had a sensory system that was ineffective part of the time.
One assumption is that the ineffectiveness of our main sensory system (vision)
at night puts us at risk from more effective night predators ("we" being
organisms without sonar or effective night vision). Most sensory systems
that work in low light or without light are better at detecting moving objects
than stationary ones. It requires either an internal energy source in the
target or an energy flood from an outside source like the sun to allow easy
pattern detection that allows a predator to identify stationary prey. So
we presume that for the most part it is a good thing for day animals to stay
still at night (and for complementary reasons for night animals like bats to
stay still during the day).
If a living control system must stay still except in an emergancy, it makes
sense for (a) the muscular output systems to be paralyzed and (b) the sensory
systems that can function under those conditions to remain functional while
the sensory systems that are minimally functional are closed off. This is one
of the charactistics of sleep. The different stages of sleep have different
kinds of paralytic effect, but I have no theory about that (nor much knowledge
of the different effects).
Even under normal daytime conditions, with all output systems functioning
perfectly, most of the potentially controllable percepts are not being
actively controlled, because of the resource limitations imposed by the
restricted degrees of freedom. So a lot of them are presumably being
pseudo-controlled through the imagination loops, in coordination with
incoming sensory data (I think it likely that the imagination predicts
and tracks percepts as part of the alerting system, but I have no explicit
model for how that happens).
When one is asleep, there are no actively controlling systems, at least not
for the external musculature, most of the time. But the imagination has no
such restrictions. Only if the imagination activates some kind of alerting
system (or if the sensory systems not closed off do so) will active control
return. And as we all know, both kinds of awakening stimulus occur. It is
not noise in itself that awakes us, but noise of prespecified patterns or
of unexpected loudness or sharpness. My almost inaudible alarm clock that
my wife does not hear beeps me awake, when the loud rumble of a passing truck
does not. She might awake to the almost inaudible cry of her baby (many
years ago), when I would not hear it even after she nudged me. And for
imagination, I think few people sleep to the end of the fall off a dream
cliff, unless it ends with them flying away.
What about the content of dreams? Most psychodynamic theories seem to deal
with the dream as a kind of problem-solving analogue, and I think the PCT
approach would reinforce this view. But in imagination, as in dreams,
impossible things are not prohibited, contradictory things do not necessarily
conflict, and the temporal coherence enforced by the real world does not
apply. So a dream can be locally coherent (that coherence is enforced by
the references supplied from relatively high-level ECSs) but globally
nonsensical, as when, for example, one walks along a sand dune into the bar
of a hotel that is/was/becomes the bridge of a ship. The high-level content
is supplied by high-level reference signals that are the outputs of ECSs
experiencing error, perhaps even at the level of principles or system concepts.
Other content is supplied in part from sensory data from sound or touch or
internal sensing systems (which are not closed off in sleep), or from mid-level
ECSs experiencing error.
Why is the content rich? Why do myth and art resonate so strongly? I would
assume that the richness depends on the widespread interconnections that
make the hierarchy a distributed system. There are lots of things going on
in there at any one time, and quite possible several different high-level
ECSs with error that produce output to drive them. The resonance of myth
and art may signal that lots of us, perhaps by the nature of the natural
and social world, experience conflicts of the same kind, and these conflicts
can be expressed in many different but equally rich ways. The actual form of
expression is a reflection of the particular conflict-induced error in the
artist/author, but it also provides the input necessary to reduce the error
in the audience, because the underlying hierarchies and "normal" error patterns
are alike in artist and audience.
I think this approach could be much improved by careful thought, observation,
and experiment, but it appeals to me very much. It makes dreams, imagination,
myth, and art a very natural consequence of life. The first two need no
explicit communication devices. They are internal, and should be shared
by most animals. The second two need communication, and in the case of myth
the communication must be linguistic. Myth I would assume to be strictly a
human structure, but I'm not at all sure about art.
There is a town in the Black Forest called Freudenstadt. It's a very ordinary
small town, and a great traffic bottleneck.
All very slimy speculative, but I like it.
Martin