. 2
( 4 .)


the brain is to enable movement, allowing the animal to locate
food, escape enemies and find mates. This requires the animal
to have sense organs to examine its environment, and muscles
to actually move its body. The brain is the link between these
two, analyzing sensory information, deciding where to move,
and controlling the muscles to carry out this action. We will
focus on the first of these tasks, understanding how the inner
reality facilitates the analysis of sensory information. While it
is possible that the inner reality is also used in determining and
controlling movement, this is much more speculative and we
will not pursue it here.
To start, look at the photograph in Fig. 8-3 for a few
moments. When done, speak a sentence or two on what this
picture is about, such as if you were briefly describing it to a
Chapter 8: The Function of the Subreality Machine 123

An old photograph. This is easily recognized as a man and
a woman standing in a laboratory, taken around 1900.

Your response is probably something such as: “This is an
old photograph of a middle-aged man and woman standing in
a laboratory, probably taken about 1900.” You might have
even recognized it as a photograph of the great scientists Pierre
and Marie Curie, famous for their work on radioactivity. You
were able to extract this key information with only a few
seconds of examination. It wasn™t even difficult; this is a task
that can be quickly carried out by any normal adult.
The Inner Light Theory of Consciousness

Now suppose that we want to build a computer to perform
this same action. That is, we want to show it a picture that it
has never seen before, and have it provide a short description of
what the picture is about. We gather together a team of
engineers and scientists that are experienced in this area, such
as connecting video cameras to computers, developing software
to recognize shapes in digitized images, and creating databases
of stored information. We describe the goal of the project to
our technical team, and ask them to give us an estimate of how
long it will take, and how much it will cost. In other words, we
want to get a general idea of how difficult this task really is.
From a technical standpoint, is this something that is relatively
easy, or is it something that is relatively hard?
After hearing our goals, most of our technical group gets up
and walks out of the room, mumbling that we have wasted their
time. The few that remain are kind enough to explain. One of
them offers, “I rate the difficulty of new projects on a scale of
1 to 10, and this one is about 100." Another tells us, “Assuming
our current rate of technological learning, this is the kind of
project we might tackle 50 to 100 years from now.” Still a third
comments, “We have all the basic tools, but the overall
complexity is just too great; it reminds me of a man holding a
brick, looking up at the great pyramids.”
The point is, the analysis of sensory data is extremely
difficult, far exceeding the capabilities of present day computer
technology. We perceive it as effortless only because this brain
activity is blocked from our conscious examination.
The primary reason that sensory analysis is difficult rests
with the data itself. The information provided by our senses is
very poor quality; it is incomplete, ambiguous, contaminated
with interference, and degraded in a variety of other ways. As
an example, when you looked at Fig. 8-3 you probably didn™t
notice anything unusual. But Fig. 8-4 points out a variety of
aspects of this picture that are difficult to reconcile with the
physical world. For instance, some of the objects merge
together without a distinct boundary between them, such as
Chapter 8: The Function of the Subreality Machine 125

Image discrepancies. Vision and the other senses provide a
poor representation of the physical world.

Pierre™s foot and the floor. Other objects have an incomplete
relationship with their surroundings, such as the dark rectangle
floating in mid air. A scratch in the photograph shows up as a
horizontal line, with no relation at all to the viewed scene.
Severe problems are created by representing the three-
dimensional setting as only a two-dimensional image. This
produces missing elements, such as Marie™s legs, Pierre™s hand,
and the back side of all the objects. It also makes whole bodies
The Inner Light Theory of Consciousness

appear as discontinuous, such as the elbow being separated from
the remainder of the arm. Further, the resulting geometric
distortion changes the shape of objects, such as the rectangular
table top appearing as a parallelogram.
Your first impression might be that the comments in Fig.
8-4 are trivial and unimportant. No so; these are problems that
present day computer scientists struggle with on a day-to-day
basis. But the human brain has already solved these problems;
it is capable of finding the relevant data in the exceedingly poor
information provided by our eyes, ears, and other sense organs.
The question is, how does the brain do it so well, and what does
this have to do with an inner reality?

Filtering versus Matching
To answer this question, let™s look at two techniques
engineers have developed to analyze poor quality data. As an
example, imagine that we want to receive a radio signal from an
orbiting satellite, as illustrated in Fig. 8-5. The signal being
transmitted is very simple, nothing but a sine wave at a constant
amplitude and frequency. This is very familiar to those who
work with electronics. If you don™t have such a background,
just look at the pictures to get an idea of what is going on. The
important point is that the signal sent by the satellite is very
smooth and regular.
In an ideal situation, the signal received on the ground
would be identical to the one being transmitted by the satellite.
Unfortunately, this is never the case when dealing with signals
that have passed through the environment. As illustrated in this
figure, the received signal is very degraded; it generally
resembles the transmitted signal, but it is very jagged and
irregular. This is the result of many different problems. For
instance, the height of the peaks may fluctuate because the
satellite is in motion, or from atmospheric turbulence. In
extreme cases, this can result in sections of the received signal
being completely missing. Another problem is interference; for
instance, our receiver might inadvertency pick up the radio
Chapter 8: The Function of the Subreality Machine 127

Passing signals through the environment. The received signal is
a poor replica of the original transmitted signal, due to noise,
interference, and similar problems.

transmission from an aircraft flying overhead. This becomes
part of the received signal, degrading our ability to detect what
is coming from the satellite. Still another problem in acquired
signals is random noise, a term scientists and engineers use to
describe a wide variety of fluctuations. This results in such
things as “snow” in television pictures and static in radio
broadcasts. Random noise can arise from many different
sources, including the mere motion of atoms and electrons. In
our example of Fig. 8-5, this type of noise shows up as a
“roughness” in the received signal.
The key point is that the signal we receive on the ground is
a poor quality replica of the signal transmitted by the satellite.
It is distorted, missing sections, and contaminated with random
noise and interference. The question is, what do we do about
The Inner Light Theory of Consciousness

it? How can we change the received signal to more resemble
the original?
Figure 8-6a shows our first approach to this problem, what
engineers call filtering. There are many different ways to carry
this out, and we will only give a general description leaving out
the technical details. The basic idea is to pass the signal
through an electronic circuit or computer routine that changes
the signal™s characteristics in some desirable way. For instance,
if we know that the signal being transmitted from the satellite is
relatively smooth, our filter might remove the roughness in the
received signal, as illustrated in this figure. If you don™t have
a background in electronics, think of this as performing the
same function as the suspension on an automobile, providing a
smooth ride even over a bumpy road. Filters are very common
in electronic circuits, and can be very simple to extremely
complex. But even the most advanced filters have limitations
on how well they can work with highly degraded data. As in
this example, when interference and random noise dominate the
received signal, the output of the filter still looks like
interference and random noise.
Now we want to turn our attention to an alternative
technique, called the phase lock loop. This is far less common
in electronics, being used in only a few specialty applications.
Just as before, we will only give a general description that
leaves out the technical details. As shown in Fig. 8-6b, the
phase lock loop is composed of two parts, a comparing circuit
and a sine wave generator. The sine wave generator does just
that; it produces a pure sine wave, without distortion,
interference or noise. The function of the comparing circuit is
to continually compare this created signal with the signal
received from the satellite. If a difference is found between the
two, the comparing circuit generates a “correction signal” that
is fed into the sine wave generator. This, in turn, causes the sine
wave generator to alter its output in an appropriate way to make
a better match. The overall effect is that the phase lock loop
generates a perfect sine wave that is the best possible match
Chapter 8: The Function of the Subreality Machine 129

Filtering and PLL operation. As illustrated in (a), filtering
attempts to “clean up” a contaminated signal. In comparison, (b)
shows how a phase lock loop generates an entirely new signal.

to the received signal. Even if the satellite stops transmitting,
the phase lock loop will still produce a pure sine wave output,
its best match to the remaining random noise and interference.
The phase lock loop has one tremendous advantage and one
tremendous disadvantage compared to filtering. The advantage
is that it can operate with extremely high levels of interference
and random noise, while still producing a near ideal output.
Filtering can™t come close to matching the phase lock loop in
this respect. The disadvantage is that the phase lock loop only
knows how to detect one very specific thing, a pure sine wave.
The Inner Light Theory of Consciousness

For instance, if the satellite started to transmit a waveform of
some other shape, the phase lock loop would respond in the
same old way, producing a sine wave output. In short, the
phase lock loop works well with degraded data, because it is
only looking for a single thing.
It is a commonplace belief that our minds directly perceive
the physical universe. As an engineer would put it, the objects
around us result in signals being passing into the brain, where
they are somehow perceived by our conscious minds. Various
filtering operations may be applied to these signals by our
neural circuits, but what we end up experiencing still has a one-
to-one correspondence with the external world. However, this
view is simply not true. The brain does not “filter” the signals;
it generates new signals that it believes are the best matches to
the nearby environment. In other words, it operates like a phase
lock loop, not an electronic filter.
As we move about the world in our day-to-day activities,
our brains must continually keep track of what is around us.
The brain is also responsible for identifying other aspects of the
local environment, such as its sounds, smells, and tastes. This
information about the surroundings comes to the brain through
the senses, usually in a highly degraded form.
The brain™s task is to extract relevant information from this
jumble of interference and noise, allowing it to plan and execute
movements. To do this, it takes advantage of the fact that
nearly everything it encounters is familiar. Our daily lives are
composed of objects and situations that we have experienced
many times before. This means that the brain does not need to
identify every possible pattern and scenario that could ever
exist. On the contrary, during most of our conscious lives our
brain only needs to recognize those things that it has recognized
in the past. Just as the phase lock loop only looks for a single
waveform, the brain only needs to look for a limited number of
patterns. That is, at least most of the time.
As a demonstration of this, look at the “ambiguous” figures
shown in Fig. 8-7. These are illustrations that can be interpreted
Chapter 8: The Function of the Subreality Machine 131

Ambiguous figures. On the left is “Rubin™s vase,” named after
Danish psychologist Edgar Rubin who first presented it in 1921.
This figure can be alternately seen as a black vase, or as two white
faces in profile. The illustration on the right is often referred to as
the “Boring figure,” after psychologist E.G. Boring who explored
the psychology of it in the 1930s. This figure can be seen as either
a young woman or an old woman. It dates to at least the 1890s,
when the Anchor Buggy company used it in an advertisement with
the caption: “You see my wife, but where is my mother-in-law?”


. 2
( 4 .)