T
he
possibility of the reflective level of neurological organization.
Having seen how a higher level of neurological
organization could open up an entire range of new powers to control
relevant conditions, we can conclude that its evolution is
inevitable, if it is possible, in the sense that it can be
tried out as a random variation and naturally selected for such a
function. There is not much reason to doubt its possibility when it
comes to the capacity of the mechanism of embryological development
to try setting up brains with reflective imagination, because the
reflective level occurs in the faculty of imagination. But at first,
the reflective level of neurological organization would be much
simpler and weaker than the capacity for reason we have been
considering, since the brain mechanisms on which beings like us can
reflect on are informed by many centuries of cultural evolution.
Thus, it may not be obvious how, even if a higher level of
neurological organization were tried out as a random variation, it
would be naturally selected. But before we take up that problem, let
us make sure that the reflective level of neurological organization
is as simple as it seems to be.
S
tructure
of the reflective brain. The functional diagram of the
reflecting animal behavior guidance system shows how nervous systems
of the kind we have been considering could incorporate a mechanism
for rational imagination. Since the higher level of neurological
organization occurs within the faculty of naturalistic imagination,
it occurs in each of the systems involved in generating verbal
behavior in mere linguistic animals. The psychological linguistic act
and psychological image are represented as more encompassing forms of
behavior than naturalistic acts and images, because the latter are
contained as special parts of the former.
Speaking.
In the speaker, the psychological linguistic schema must not only
construct all the images of psychological states (beliefs and
desires, including both the proposition and the propositional
attitude toward it in each case), but also predicate the
psychological states of an animal subject. In order to express this
nonverbal activity in the rational imagination verbally, a more
complex grammatical structure is required. The sentence must identify
the naturalistic image (or psychological image) whose causal role is
being described, it must identify the kind of causal role predicated
of it, and it must identify the subject in whom such a psychological
state occurs. Since there is already one predication in any
naturalistic image toward which the propositional attitude is taken,
a basically new kind of grammatical marker is required to indicate
the kind of predication that occurs in constructing psychological
images. Those grammatical markers are the verbs of propositional
attitude.
Verbs
of propositional attitude indicate a kind of predication in which a
naturalistic (or psychological) image is represented as playing
certain kinds of causal roles in causing behavior (or beliefs). That
is, psychological linguistic schemata must be able to construct
naturalistic (or other psychological) images in the sensory input
system and then temporarily impose them on the speakers brain so that
their effects on the brain can represent their implications about
intentions (or beliefs). The verbs of propositional attitude
represent how they are imposed on the brain, and given the nature of
animal behavior guidance systems, there are three basically different
causal roles they might have.
Verbs of
propositional attitude such as “believe,” “perceive,” and
“remember” indicate various ways of imposing naturalistic (or
psychological) images on the sensory input system. They constrain the
speaker’s behavior guiding processes so that the sequences of
images that are called up from memory in rational imagination
represent what a subject would infer in certain situations.
Verbs of
propositional attitude such as “desires,” “needs,” and
“wants” indicate the imposition of naturalistic (psychological)
images on the goal direction system of the speaker’s brain. They
constrain the speaker’s brain so that the sequences of images that
are called up from memory represent what a subject would try to do
when certain kinds of behavioral schemata were activated (relative to
certain objects).
Verbs of
propositional attitude representing intentions indicate the
imposition of naturalistic (or psychological) images on the speaker’s
behavioral output system. They constrain the speaker’s brain so
that the sequences of images that are called up in rational
imagination represent the beliefs or desires that might lead a
subject to behave with such an intention.
Though
there are three basically different kinds of causal roles that such
first level propositions might play in behavior guidance processes,
the reflective linguistic schemata need only two basically different
ways of imposing propositions on the brain, because desires and
intentions are both represented in the behavioral output system.
The goal
selection system does not represent goals on its own, but only by
activating certain kinds of behavioral schemata in the behavioral
output system (perhaps, relative to certain objects in the sensory
input system). Thus, desires can be represented in rational
imagination by imposing propositions on the behavioral output system,
for if the kinds of behavior represented are general enough in the
situation, the sequences of images called up from memory will be
inferences about how to behave in that way (perhaps relative to so
object picked out in the sensory input system)
Since
there are only two basically different verbs of propositional
attitude, no great changes are required of biological evolution in
order for linguistic brains to acquire the ability to use
psychological sentences. It would be enough for some random variation
to try out a mechanism for imposing propositions (naturalistic or
psychological images) on the speakers brain in two different ways,
one affecting mainly the sensory input system and the other affecting
mainly the behavioral output system. The distinctions among kinds of
psychological states is each class could then be acquired by
learning.
For
example, the difference between imposing a naturalistic image as a
perception or imposing it as a belief could be acquired as part of
the process of learning to impose propositions on the sensory input
system, for the difference would show up mainly in the strength of
the belief and how easily it could be discounted.
Likewise,
the difference between imposing a proposition as a desire and
imposing it as an intention could be acquired as a refinement of the
capacity to impose propositions on the behavioral output system, for
it would merely indicate which of the sequences of images that were
called up from memory in rational imagination is relevant, those
representing inferences about how to behave (that is, its
consequences) and those representing inferences about the beliefs and
desires that motivate it (or its causes).
In
order for verbal behavior to indicate the activity in the speaker’s
rational imagination, the nonverbal covert side of the linguistic act
must be expressed verbally, and the job of representing the activity
in rational imagination in verbal behavior is performed by Wernicke’s
and Broca’s areas as indicated in the functional diagram of the
reflective brain. (The projection from the former to the latter is
not represented in this functional diagram, but it can be seen in
(See the diagram of the linguistic brain in the Linguistic
stage, Stage 8.)
The
motor output for speaking comes from Broca’s area. But Broca’s
area depends on input from Wernicke’s area, because Wernicke’s
area is what receives information from rational imagination about how
the psychological image is being constructed.
Wernicke’s
area must not only give the names of the images (meanings of words)
involved, but also have grammatical markers that distinguish one kind
of predication from another (that is, the naturalistic predication
that may occur in the embedded naturalistic sentence from the
predication of the psychological state to some subject). That is the
role of verbs of propositional attitude: they are the grammatical
markers for the new kind of predication, the one that is involved in
imagining subjects as having psychological states.
Since
grammatical markers for the predication involved in the construction
of naturalistic images has already evolved, they can be embedded in
the psychological sentence after the verb of propositional attitude.
That is, the naturalistic sentence is contained in the psychological
sentence after the verb of propositional attitude, like a name for
the naturalistic image toward which the attitude is taken.
With the
capacity to distinguish psychological from naturalistic predication,
it is possible to treat psychological images in the same way as
naturalistic images in rational imagination. Thus, subjects can be
seen as having beliefs or desires about beliefs or desires.
Assuming
that Wernicke’s area supplies the required words and grammatical
markers, Broca’s area can generate the motor commands for the words
and the grammatical markers through the motor neocortex. Once again,
verbal behavior would be generated by the linguistic schema only
indirectly, as a result of thinking the content of the sentence and
the effect of the images constructed in imagination on Broca’s
area. (See Linguistic stage:
Possibility: Structure for details.
Though only
the verbal side of linguistic behavior is overt, it can also be
generated covertly. That is, one can speak to oneself silently by
generating the motor commands for speaking covertly. But it is also
possible to carry out the nonverbal side of a linguistic act without
generating even covert verbal behavior. That is how the reflective
subject normally thinks.
Listening.
The listener’s understanding of psychological sentences parallels
that of natural sentences. Wernicke’s area constructs the
psychological image in the listener’s rational imagination using
telesensory input about the speaker’s overt verbal behavior.
Wernicke’s
area acquires the nervous mechanisms for decoding verbal behavior in
the development of the capacity to speak psychological sentences. The
connections between words and the images that are their meanings and
between grammatical markers and the ways of combining images that are
their meanings are two-way connections (as indicated in the
functional diagram). That is, Wernicke’s area uses the words
identified in telesensory input to call up the images and the
grammatical markers identified in telesensory input to combine them
on the inferior parietal role in the same way that the posterior
cingulate neocortex would have, if the brain had been speaking the
sentence meaningfully. Thus, the listener can grasp the meaning of
the sentence heard. Broca’s area is not required for comprehension.
Unlike
the previous levels of neurological organization, I cannot point to
any additional brain structures that realize the higher level of
neurological organization. Its existence is basically an inference
from the structure of psychological sentences and their function as a
faculty of rational imagination. But given the basic structure of the
faculty of imagination, it is plausible, at least, to suppose that
there is some new brain structures that enables the reflective brain
to impose propositions on the brain in the two basically different
ways described above.
Wernicke’s
and Broca’s areas of the neocortex are properly evidence of
reflective level neurological mechanisms. I cited them as evidence
for the evolution of the mechanisms required for the use of natural
sentence, but that is just an inference. There are no brains left
from the primitive spiritual stage of evolution to show that these
brain mechanisms evolved first in linguistic brains. Even if there
were, there might be no dramatic evidence of the difference between
these two levels of neurological organization, because the reflective
level of part-whole complexity does not require any basically new
systems. It requires only higher levels of part-whole complexity in
the same systems that constitute naturalistic imagination with its
capacity to generate and be controlled by over verbal behavior.
There
may, however, be some evidence for this theory about the difference
between beliefs and desires in the way different kinds of words are
processed in Wernicke’s area. Different regions of neocortex in
Wernicke’s area are the locus for words referring to objects in
space, such as kinds of animals, and words referring to kinds of
behavior, including tools. It suggest that there is a connection
between telesensory images of each kind of behavior and behavioral
schemata for generating that kind of behavior. This is evidence of a
mechanism by which the brain distinguishes between beliefs and
desires, that is, between psychological states most relevant to input
and those most relevant to output.
There
is little danger in proceeding at this point on the assumption that
there is a reflective level of neurological organization that can be
tried out as a random variation. Given the structure of the faculty
of imagination, only rather modest changes are required for the
behavior generator to speak psychological sentences and, once again,
the details of the nervous mechanism can be acquired by learning.