"Men believe themselves to be free, simply because they are conscious of their actions, and unconscious of the causes whereby those actions are determined." Espinoza, 1955
Are they always ‘talking’ to each other?
The boast for the rational architecture of reality as proposed by brain scientists is predicated on the belief that their conclusions rest on the reliable scaffolding of reason and the ‘logical’ inferences from known ‘facts’. They often seem to forget that their systematic descriptions are not absolute, objective assertions about objects or events in nature but merely symbolic representations of low-fidelity sensory intuitions encoded in a system of signs and rules for their use, something we call language. In a previous chapter “Visceral Brain, Thoughts and Language.” we warned about ‘non-logical contaminants’ when we concluded that “..language development in the form of speech and writing are not fully developed when the same anatomical structures in their production had been in use to establish stereotyped responses to manage the emotional, hormonal and nutritional life-saving needs early on and these would now become inextricably commingled with their new use in speech and sound output. If language is anywhere involved in the generation of thought, as we have suggested here, this situation becomes a most fundamental limitation for man to ascertain the true reality in itself of objects externally perceived.
To establish the ubiquitous presence of a biological survival primacy as over-riding another biological imperative to 'logically' (but non-inferentially) structure the unlabeled world out there we offered into evidence various behavioral and neuro-pathological data like e.g., the developing toddler’s first choice of syntax as being attuned more to the functionality of their survival effort than the acquisition of ‘syntactic correctness’ of their parent’s language, e.g., “baby eat”, subject and predicate. We argued this observation as representing the unconscious choice of understanding (semantic consideration) the hierarchical primacy of the hunger problem to be solved. The evoked linguistic response is more adequate to the peremptory solution (reduction or elimination of urge) than the elaboration of a syntactic ordering of a vocalization. It is upon these innate semantic language structures (proto-semantic referentials) that the developing newborn adds on a 'form' by acquiring a particular language from their parents and refine the pertinent proto-semantics by giving it a propositional architecture (inferential syntax) as time goes on. This can only happen if the semantic component development (angular gyrus?) over-rides the controls for the linguistic production at this early stage. One can suppose that any acquired syntactic language corrections at this stage may be considered adaptively inefficient in the communication of basic biological needs to the mother. This primitive language structure can be recalled in adults by hypnosis. There is no doubt that the child’s insistence on linguistic shortcuts before parental enforcement of a particular language structure responds to an inherited 'logic' (but not inferential) predisposition tied up with survival of the individual species. This implies that the multi-layered neocortical arrangement, ideal for the parallel processing of external sensory input later on in development, is being used at this early stage (along with paleocortical neuronal circuitry) to attend the survival-oriented, stereotyped requirements that precede the cephalization of human functions. At the pre-linguistic post-natal stage, judging from the behavioral data, primitive 'syntax', disorganized babbling, startle responses and uncoordinated limb movements do not necessarily require a strategic command center, but rather an interactive cooperative effort between sensory and motor components of a very primitive reflex arc response incorporating visceral and emotional stereotyped responses to the internal threshold body physiology variations.
By extending the scope of interaction between the competing needs for biological survival and logical structuring of the external world, we now go beyond the visceral brain ‘contamination’ to integrate now elements of the internal and external environment and observe how they get commingled within the context of the pleasure / aversive emotions they generate, code for and store in the process. It is contingent upon the subsequent attainment of the relevant intellectual, emotional, pre / post linguistic and psychological structures, that the concept of an (external) object in the environment slowly emerges in the newborn. From then on, the objects have an existence and permanence of their own. This unconscious awareness signal heralds the dawning of a primitive conscience of self.
We have detailed earlier important neural input connections from viscera, taste receptors and peripheral nerves, now it is time to focus briefly on the olfactory cortex, limbic system (medial forebrain bundle), amygdaloidal body, hippocampus and anterior / medial thalamus and the hypothalamus as they play significant roles in the genesis of the unconscious, psychic and motor correlates of the emotions. We will argue how the conscious ‘feeling’ of these emotions (qualia) may require extra-physical explanations and form an essential, inseparable component of our rational formulations of nature. In another chapter we will develop the proposition that, while we may be able to 'unconsciously' behave under the influence of an emotion, we can only 'consciously' interpret it when we access higher order system that makes possible the conscious experience or feeling of qualia.
The Lamarckian evolution ‘heresy’ revisited
It has been a long time since the undersigned debated his teachers as a pre-teen on the self-evident logic of inheritance of environmental traits. How else could one explain the capability of organisms to display adaptive responses to an environment no longer present? We even set out to explore in the laboratory for mechanisms of information transfer from genetically altered somatic cells (or labeled fragments thereof) to gametocytes in an effort to circumvent the Weissman’s dogma on the ‘continuity of germinal plasma’. We could not do it experimentally decades back before the molecular genetics revolution, so we suggested that evolutionary theory may have to consider alternative methods of selection, why not add to the seemingly random occurrence of environmental selection, a conscious and deliberate selection of a sexual mate to achieve what blind evolution couldn't possibly explain, a novel way of accelerating Darwinian evolution! This was left for future generations to work out the details. Recently, the Human Genome Project demonstrated how the bulk of DNA encodes various aspects of environmental traits that can be directly inserted into a newborn at the moment of fertilization. One can speculate that environmental information carrying survival alternatives for the species is somewhat woven into the classical metabolism-oriented information we associate with the genetic code, there is plenty of 'nonsense' code in DNA to be explained. Lamarcque and Lysenko couldn't have imagined! The interactive correlation between these two distinct forms of evolutionary selection processes, the seemingly random Darwinian selection and the deliberate, sexual selection processes has to be worked out. We suspect that it may hold the key to understand how the off-line DNA-coded genetic memory information interacts with the on-line social information that our sensory receptors are continually feeding as input into the system. No doubt that the genetic code we inherited already contained information about the ecological / social? environment we are likely to encounter when born; how do we incorporate survival-relevant environmental information into a heritable DNA code still has to be worked out. How else can we explain the layout of the developing nervous system so suited to the exigent demands of the ecosystem we are likely to be born into, as we shall soon learn.
When we examine the longitudinal organization of the central nervous system (cns) it is noticed a centrifugal layering of neuronal tissue about the central neuraxis. The most internal archilayers (hereinafter to be referred to as Module 1) have the most primitive conservative functions, are redundant, are the first ones to recover from injury and tend to form a reticular array of collaterals at all levels (reticular formation). At the spinal level, they coordinate the motor reflex activity of the newborn, at the cerebral trunk they coordinate the activity of the vital cardio-respiratory, deglutition and emetic centers, among others. At levels higher than the pons, they organize the reticular activating system (ras) and their ascending and descending projections and receive synaptic connections from all sensory information channel collaterals (except olfaction). This information they relay to the cortex (cortical alert) via the thalamus. The intermediate paleolayer (hereinafter to be referred to as Module 2) houses the sensorimotor information channels. The external neolayer (hereinafter to be referred to as Module 3) of most recent evolutionary acquisition becomes functional by a progressive myelinization process starting ca. 2 years of age and being completed at ca. 5 years old when required for fine dactilary movements. Its superficial position makes it more vulnerable to trauma without the benefit of the redundancy seen in the deeper reticular archilayers. (See Vol. 1, Human Biology, MightyWords.com). It will not be an exaggeration to generalize and conclude that the primitive internal (medial) archilayers are the custodian of vital evolutionary genetic information and are charged with the awesome responsibility of preserving the anatomico-physiological integrity of the species before the ever-changing manifestations from a hostile physical and mental environment. We will now extend the generalization to include important observations: they process information in non-linear patterns and thus their effects do not have access to consciousness via the ‘talking’ dominant brain, preferring instead the utilization of representational, non-linguistic conduits to consciousness in the form of an emotional qualia, an unconscious mental state not requiring an interpretation of the environmental situation yet. Finally, we will speculate, and elaborate further on, about its possible role in the control of our thoughts and rational lives as well.
Underlying unconscious and irrational vital activities
Digital technology has now made it possible to measure with precision the conduction time along neural pathways as a function of synaptic delays encountered and the diameter of the conducting fibers engaged. We can also map the geography of the conducting pathway by using specialized neuro-anatomical techniques, e.g., peroxidase distribution. We can best illustrate the usefulness of these lab techniques in the demonstration of these parameters by giving an example. Florida is a haven for all kinds of crawling, jumping or flying critters with their wide assortment of behavioral displays covering the entire audiovisual spectrum. After being stung by a fire ant or a yellow jacket or idly chatting within striking distance of an unsuspected rattler or a black bear, you ‘learn’ pretty fast to distinguish their distinctive wavelengths of communication or do you just ‘remember’ them? Let us examine some recorded encounters. A tourist visiting from N.Y. City is camping inside one of many camping sites and wanders, distracted by the colorful contrasts, into the wilderness when surprised by the sudden rustling noise of the nearby tall underbrush. The ‘unfamiliar’ noise represents a stimulus that has been modified or conditioned to unleash a wide display of bodily responses in anticipation of an impending attack the noise is warning him about. He freezes in his tracks quicker than a cat can wink its eye, ready to run away or attack. The entire spectrum of sympathetic autonomic defensive responses are mobilized as described by Cannon, including cardiovascular, respiratory, muscular and neuro-humoral alert adjustments to meet the unknown, even the secretion of analgesic endorphins to tolerate the pain anticipated from the impending attack. The freezing response is so fast that one can easily rule out conscious processes initially participating. Similar situations have been simulated in various laboratories with different species of animals. The sensory arm of the complex reflex response follows two different audio-visual pathways en route to their corresponding memory data banks. In a previous communication, aforementioned, we have detailed the auditory pathway, contra lateral to the entry of the sound stimulation en route to higher levels of differential analysis in the auditory cortex > hippocampus of the temporal lobe, the ‘explicit’ memory site we will return to later on. Before that, let us trace the faster pathway to the ‘implicit’ memory storage site, the amygdala, a small sub cortical forebrain area sometimes confused with ‘short term’ memory site.
Joseph Ledoux has traced the central sensory acoustic fibers from the ipsilateral cochlear nucleus to the contra lateral inferior colliculus of the mesencephalon > medial geniculate body of the thalamus > primary auditory cortex + lateral amygdala > central amygdala. (see The Emotional Brain, 1996 Simon & Schuster). The central nucleus of the amygdala is now the command center coordinating and completing the fast survival avoidance motor response reflex arc by connecting with the various autonomic centers in the brain stem associated with the Cannon response. Bruce Kapp also demonstrated a ‘freezing response’ after stimulating the central nucleus. Others have shown connections with the mesencephalic periaqueductal grey where endorphins are also produced and the hypothalamus and stria terminalis, both associated with the release of stress hormones. Connections with basal ganglia controlling postural skeletal muscle responses are not that clear yet.
This is not the description of just another super fast avoidance reflex arc. The central amygdala probably represents the reservoir of inherited information, genetic memory, specializing in the biological survival of the human species. Continues below -->
Its output depends and gets its cues from an audio-visual or multimodal gallery of recorded primitive responses all geared to protect the species from a repertoire of dangers, real or no longer present, except for their multimodal signature sensed. Meanwhile, the higher processing systems in the neocortex are busy in their processing of the same sensory information trying to figure out, what is this rustle sound and brush movement all about (context analysis)? Once Module 1 'archilayers' have concluded their fast emergency adjustment (freeze reaction) they must wait for further instructions from the slower multimodal parallel processors in Module 2 'paleolayers'. Meanwhile, the hypothalamus elaborates the intermediate neurohumoral motor and affective responses. This is achieved by its connections to the prefrontal cortex directly and by way of medial thalamic nuclei, this way providing the pathway for the genesis of unconscious mental states associated with visceral functions. During all of this increased cephalization of the processing the hippocampus sends connections to the parietal angular gyrus which itself receives information from the visual and auditory areas of Wernicke (Barr’s The Human Nervous System, Harper Row, 1975). Module 2 output report has linear and non-linear elements, as we will briefly examine below. Once the 'context' analysis is ready it is possible that upon the central amygdala getting this report, there will be a relaxation from the original motor inhibition response as the probabilities of those sounds originating from a dangerous encounter diminish. The important thing to keep in mind is that during this brief succession of 'analytical' (but non-inferential) processing events the subject is unable to articulate his stereotyped responses in a narrative because all of his responses at this stage have been servo-controlled, unconscious processes. It is during this mental stage of the process that an unexplainable, un-articulated qualia of fear and expectation anxiety drowns reason but does not exactly disappear completely as the ‘talking’ brain, Module 3, tries to reason out the occurrence and articulates an explanation, audible or not (inner speech). This view harmonizes the famous William James-Cannon controversy as to whether we experience fear and anxiety (without being able to understand and explain it) before we freeze or after that motor inhibitory response? We believe that the quicker ‘freeze response’ was controlled by the content in the ‘implicit’ memory of the amygdala while the unconscious un-articulated fear quale is controlled by the hippocampus (links to affect networks) while the 'context' is being 'analyzed' (non-inferentially?). The hippocampus formation is the most likely site for the elaboration and temporary storage of ‘long term’ memory which later on is transferred to the pre-frontal executive cortex. Needless to say that soon after our tourist sees his host’s retriever dog emerging from behind the underbrush, she sighs, her homeostatic controls in Modules 1 & 2 have now taken over as she exclaims, ”..how stupid of me.., this park is all fenced in and full of park guards!” This is a typical post-facto rational assessment articulated by Module 3, the ‘talking brain’ under Module 3 'affective' (conscious) influence (as opposed to Module 1 raw, unexplained 'emotional' background). In reality there always remains a background level of alertness, the workings of the reticular activating system (ras of Module 1) maintaining both the sensory thalamo-cortical relay pathways and the amygdala in ‘standby alert status’, as recorded. This late lingering reaction may be explained by the ‘environmental context’ (wilderness, underbrush) in which the conditioning stimulus (rustling of leaves) took place.
This assessment that differentiates between two different unconscious-level subsets of affective mental states needs clarification. Language habits limits a clear distinction between an un-explained emotion of fear (amygdala-controlled reticular archilayers of Module 1) followed by a nagging 'sense' (conscious) of emotional apprehension following the 'context' analysis (hippocampus-controlled paleolayers of Module 2) while the neuro-hormonal preparation for a contingent Cannon response was taking place also at Module 1. Each subset represents a graded increase in the level of consciousness. A transition from unconscious awareness --> to subconscious awareness has taken place each associated with its characteristic granularity of qualic sensations. Should the environmental stimulus (external or body proper) be novel then the agent has to appeal (access) to higher order inferential networks requiring the subject-agent becoming simultaneously the object of an introspective analysis. The subject becomes the 'I' and the 'subject-object' becomes the 'you' comunicating through the aegis of an 'inner speech' soliloquium. In summary, genetic-controlled awareness --> genetic / acquired controlled consciousness --> introspective self-consciousness. We will argue that the latter is a specific result of the emergence of the cortical neolayers that made possible the inherited proto-semantic information potential be modified and integrated with the acquired environmental information about objects, their attributes and inter relations as structured by the evolving language adopted and culminating in the higher order thought. If necessary it will be causally efficient in the execution of an adaptive response to exigent circumstances that continuously challenge the species biopsychosociological (bps) survival potential.
Can this example of ‘context conditioning’ and the preceding avoidance reflex response to an absent danger encountered by same species eons ago qualify for a Lamarckian inheritance status? We suspect so. If not, how else can one explain the following experiment? We know that destroying the amygdala we can inhibit the unconscious conditioned responses we just described but we can still train an animal to respond reflexly with fear to the presentation of a neutral conditioned stimulus (e.g., a noise) before the appearance of the nociceptive unconditioned stimulation (e.g., electric shock) as long as the hippocampus is patent, thereby demonstrating its participation in cognitive processes. In the example above the NY tourist unconsciously responded to a neutral stimulus (rustling noise) in anticipation of a potential danger it had never experienced before by executing a survival script coded in his inherited amygdaloidal implicit memory. Previous generational encounters become eventually encoded into the archilayers of the amygdala and paleolayers of the hippocampus to be genetically transmitted (via non-sense DNA segments) to future generations. The temporal sequence of responses from the unconscious quick avoidance reflexes to the slower sensory re-appraisal of the environmental context scenario, i.e., that transition from an intermediate protective affective anxiety to the the calm, conscious, verbalized rationalization that followed supposes a most complex balancing act coupling the quick, unconscious, non-linear processing of Module 1, the sensory, unconscious, multimodal re-assessment, thalamo-cortical parallel processing of Module 2 described and the conscious, language controlled, linear processing of all inputs coming serially into the propositional Module 3 which now must assure harmony between the ongoing environmental scenario and the preservation of self, physically and mentally. This schemata will be developed further on and in subsequent chapters.
The translation of the unconscious, irrational code input into propositions
Before we proceed it must be clear that audio-visual information arriving from environmental sensors have 3 different ways of reaching an effector neuronal network able to elaborate and execute an adaptive response (see Fig. 1, below). Once the sensory information reaches the corresponding thalamic nuclei (geniculate bodies) fast route 1 will bring the information directly to the amygdala as described above; slower route 2 will bring it to the hippocampus from Wernicke’s area by-passing angular gyrus; slowest route 3 will bring it to the angular gyrus and the follow the route 5A > 3 (3’). It is worthwhile to back track a little to our previous discussion of ‘context conditioning’. At that time we stressed the importance of the hippocampus in the elicitation of this response (specifically lesioned by Herpes Encephalitis), the amygdala alone was not able to capture the importance of the ‘environmental context’ within which the conditioning stimulus (rustling noise) takes place, the ‘implicit memory’ could only decode the noise or some primitive visual cue (motion). What is the difference between both brain structures in the handling of environmental perceptions? The difference is that the hippocampus does not receive the environmental information directly from the thalamic relay nuclei like the amygdala, that audio-visual information has first to reach temporal lobe (thalamo cortical relay) and be subjected to primary and secondary cortical (transitional) processing before it reaches the hippocampus (route 2). We can say that after the amygdala concluded its primitive ‘perceptual’ categorizations it was ready to execute an adaptive response thus by-passing the longer pathway associated with ‘pre-conceptual categorizations’ in the intervening ‘transitional’ cortical areas (entorhinal, perirhinal, parahippocampal cortex) en route to the hippocampus. Notice how primitive multimodal components of the context scenery are present and able to incorporate second level post-perceptual (or pre-conceptual?) components in eliciting a second wave of arousal (ras) and affective responses (cingulate cortex) but still unable to attain levels of consciousness. Why?, because what is missing from the ‘environmental context’ code in routes 1 & 2 is the semantic and lexical component present in route 3 or phonological route (Wernicke + hippocampus > angular gyrus > 5a > 3) described in the diagram below, Fig. 1.
The hippocampal route 2, while containing the input contributions from primary / secondary sensory cortex, had to be funneled into the transitional cortex for subsequent primitive, unconscious ‘pre-conceptual’ categorizations, the anlage of conscious linguistic linear processing present in route 3. That primitive sound code gets recycled between the hippocampus and the amygdala and gets integrated into the ‘context presentation’ and, in our opinion, may well represent the ‘binding’ factor necessary for the linear processing along route 3 of the integrated constitutive multimodal elements by Module 3. Once the sentential or equivalent symbolic coding gets linearly processed, the context environmental scenery will come alive in living colors and maintained active by reverberation or re-entrant loops sustained by a repetition of the conditioned noise (or structured word equivalent), audibly or not.Of course we are describing a first level judgment consciousness (awareness) accomplished by route 3 but we cannot explain it by the analysis that preceded because nobody can find a brain mechanism to explain the ‘qualia’ that is inseparable from human awareness. We have just described a silicon brain awareness. Heroic attempts to describe higher level judgments (meta-consciousness), e.g., Darwinian model of Edelman is essentially another functionalist description of a silicon brain awareness. (See our Amazon.com review of his book ‘Remembered Present’). We will examine some of his premises below.
Fig. 1 (Legend)
Phonological Routes. (upper left) Tested by reading aloud, even words not understood (semantic deficiency) or not pronounceable (syntactic deficiency). Other deficiencies are: anomia where subjects can not find names for things, or can not match spoken words with pictures or produce speech spontaneously .
Tests (where vision and audition are normal):
A. Spontaneous talking (Route 8>>6)
Only with syntax (3>>6)
Only with meaning (2>>6)
B. Repeating sounds (1>>6)
Only with syntax (5>3>>6)
Only with meaning (2>>6)
C. Reading loud (4>>7)
Only with syntax (5’>3>>7)
Only with meaning (2>>7)
Lexical Routes. (upper right) Requires prior convergence (8) in a putative frontal cortex where Motor image is assembled (Broca’s?)
A) Spontaneous writing (8>>7’)
Only with syntax (Memory>Visual image>5’>3’>>7’)
Only with meaning (Memory>Angular gyrus>>2’>>7’)
B) Copying from text (1’>>7’)
Only with syntax (5’>3’>>7’)
Only with meaning (2’>>7’)
A) Writing from dictation (4’>>7’)
Only with syntax (5>3’>>7’)
Only with meaning (2’>>7’)
Anthropo-logic and eco-logic solutions.
The logical integration of the preceding brain events under normal conditions where the homeostatic / emotional, and perceptual parameters are factored in while in a state of hysteresis, i.e., reaching a dynamic equilibrium state, is the function of Module 3. This idea can be subjected to further analysis for what it really represents, not really a model explaining subjective conscious experience, but rather a silicon brain attempt to account for observed reality according to the combined AIM three-component vector space ("conscious state space") comprising one's Activation state (high-low), Information source (sensory external-internal) and Mode of processing (cholinergic-aminergic neurochemical modulation).
The required output syllogism from Module 3 will vary according to the premises being supplied by the other 2 modules. For example, lesions to the amygdala (Kluver-Bucy syndrome) would have caused the rustling sounds of the leaves and the motion derived therefrom as not representing a threat to the tourist of the example (e.g., unable to match the sensory environmental elements with the ‘gallery’ of danger sounds evolution encoded in the data base). The central nucleus cannot now elaborate an appropriate unconscious reflex withdrawal response initially nor assume a subsequent conscious avoidance attitude because Module 3 has rationally concluded it is safe to examine what is causing the strange sounds and movements in the underbrush, based on the information input from the other modules. The anthropo-logic conclusion would result in an ecological disaster when the hungry black bear emerges from the brush. In addition, of course, it wouldn't have been a 'logical' conclusion had the facts presented for linear processing reflected the reality of the perception. But, the deficiency was operating at unconscious levels and thus inaccessible for processing by Module 3. In other words, the logic equation depends on the truth of the premises being considered. Some other times Module 3 will ‘confabulate’ and supply a ‘fact’ not present in the environmental sensory reality; e.g., our visual cortex does not have the resolution to detect the motion of the hour hand in a clock yet we ‘logically’ infer from two different positions in space as a function of time that they ‘moved’.
Most psychology students remember the case of Dr. Claparade’s amnesic patient who needed to shake hands and be re-introduced to the therapist every time he left the room and returned during a treatment session. Her hippocampal lesion left the amygdala intact. The therapist decided then to hide a thumbtack in his hand while extending his arm for a handshake. Soon thereafter, the patient still did not remember the therapist but would not shake his hands upon his return either. When asked for an explanation for her refusal she denied knowing the reason. The therapist’s arm extension had unconsciously conditioned her to expect the sting punishment immediately thereafter and she showed fear as the arm extended (the amygdala had incorporated the arm motion conditioned stimulus into her gallery of dangerous motions and elaborated an avoidance response). However couldn’t form an adequate longer ‘working memory’ that usually requires a hippocampus, especially to relay its output into Module 3. Consequently, she consciously ‘confabulated’ and reasoned out it was ‘safe’ to receive her treatments with the ‘new’ therapist being introduced (who was causing her pain) but would not shake hands for some ‘unknown’ reason. It is not clear whether she experienced an unexplained feeling of fear when the cs was presented.
We can continue to quote from the psychiatric literature numerous examples of bizarre conscious logic reasoning (e.g., hemineglect patients -damaged right striate cortex- trying to deny her left arm exists!) based on the ‘talking’ brain processing faulty unconscious input premises arising from Module 2 misprocessing of visual information. Other times a ‘prosopagnosic’ witness will logically and honestly testify never having seen an accused perpetrator he sees every day across his fence!
Rational decisions were in all previous cases logical solutions based on false premises that did not correspond to environmental reality, far from being adaptive solutions, albeit being the result of a logical processing! Every logical conclusion we consciously reach whether we are disciplining our teenage son or proposing a hypothesis or theorem is premised on the assumption on the healthy viability of the other modular brain structures feeding information into the logic processor, nothing farther from the truth in reality. One particular case that calls our attention is that of a ‘split brain’ patient presented with an object to his Module 2 right hemisphere using special optics. When asked if he has ever seen the object he denies it because his left ‘talking’ brain did not see the object and is making the declarative statement but when blindfolded and asked to pick with his hands the same object from an assortment of different geometrical objects, he picks the correct one. It seems obvious that some memory intervened and a cognitive level discrimination was in place but there is no consciousness involved because there is no qualia, we like to think of this episode as being equivalent to the machine awareness of a silicon brain, similar to human awareness or first order judgment and far from being meta-consciousness or higher order judgment. The following comments will provide more insight into our conclusions.
Another physicalist attempt to harness the elusive mind **
Being trained as scientists one would tend to cast a consciousness model around the possibilities of an empirical demonstration in the laboratory. However, if this approach becomes a straight jacket then options for perhaps a metaphysical demonstration, theological or otherwise, should remain open, honesty requires it. This is the case here because this cannot be a theory of consciousness (considered as a second order judgment or meta-consciousness); it may be, we hope, only a credible theory of ‘silicon brain’ awareness (considered as a first order judgment). Unlike our modest brain-storming around verifiable living experiments, in the persuasive Darwinian model proposed by Edelman, et al, the authors relied heavily on the use of measurements in the EEG and artificial intelligence (AI) simulation laboratories and extrapolated their complicated traces to their origins from real or putative subjacent brain structures under varying experimental conditions. Then found the best functional model theory that fits the data. In our opinion, the 'Darwinian' model says much more than the data suggests. It should have explained clearly how three different mental stages might be coupled within the context of his "..re-entrant global mappings" to bring about the solution of the 'binding problem' that integrates imagery with qualia. In our opinion, it did not do that in a convincing way.
For instance, there is a characteristic restful, synchronous slow-wave (delta) dreamless sleep that would correspond to our Module 1 (archilayer's visceral brain acting as pacemaker, no exteroceptive, language processing or thalamo-cortical inputs); coma and anesthesia may, however, give a similar pattern. Then there is the asynchrony of REM sleep (alpha waves resembling wakefulness) that incorporates a fragmented replay (memory) of previous multimodal parallel processings of the non-dominant hemisphere (Module 2). Notice that behaviorally, both coordinated motor (muscle tone) and language processing are disabled. Then comes the asynchronous waking pattern that has now, in addition, incorporated the serial processings of the 'talking' brain (Module 3) and the thalamo-cortical reciprocal connectivity. Any theoretical elaboration of models of brain function has to be consistent with this empirical data. Edelman’s nested network approach of integrated 'functional clusters', while excellent for a possible model of machine (robotic) awareness (sans qualia) fell short of a credible model of "high order consciousness."
One can argue that a claim for a comprehensive theory must consider other types of data (behavioral) suggesting that man is a social animal whose physical integrity is a preparation to both survive (viability) and guarantee the perpetuation of his species (mating & reproduction). Man has also historically demonstrated an active curiosity to explain his origins and his destiny. To achieve the personal and social ends his activities must be conservative (adaptive) and communicated (language controlled). Edelman, et al (and others before them) have always summarily dismissed the role of language in the enactment of consciousness while timidly accepting its involvement in higher orders of activity. They take for granted that the central autonomic network organization (visceral brain) is assembled sua sponte in the newborn, no mention is made about the codified instructions for the network architecture has to be coded into the DNA. Neither is a strategy for the integration of the following functional clusters provided: Module 1, the visceral brain specializing in interoceptive monitoring, guarantees homeostatic control of vital biological survival variables by neural / humoral strategies. Module 2, the non-dominant brain specializing in exteroceptive monitoring alerts against environmental variables affecting biological survival and incorporates the multimodal manifestations of the 'un-labeled' world out there. Besides its biological survival value for the individual is also important in mating and reproductive behavior of social value. Module 3, the 'talking' brain tries to make sense of it all by processing the serially coded bi-modular input to articulate efficient individual and social strategies for the individual survival and social adaptation, the emergence of the 'self' (planned by the executive brain in the frontal lobe). The main problem with their view is that it presupposes an optimal equilibrium level between the 3 modules for maximal efficiency in the pursuit of those stated goals. Part of this problem is their implied successful integration of 3 different ways of processing information, especially when neuro-histology and electroencephalography strongly suggest a 'creative chaos' in the structural / functional architecture for the brain! A nested network theory seems like a good candidate to encompass the activity of the 3 modules especially when one considers the application of the principles of statistical entropy to 'predict' the most likely neural pathway configuration that results in effective adaptive behavior. The first big problem resides in coupling the informational output code strategies of the conservative, relatively uncomplicated Module 1 (whose main assignment is to maintain the biological status quo to execute the genetic memory instructions it inherited) with the linear processing capabilities of Module 3. Ever since our publication in 1985 of Biopsychosociology (Limusa Ed.) we have been urging the incorporation of the DNA / RNA coding as integral part of any mind / brain model. Having ignored it is in our opinion the biggest flaw in the otherwise very attractive Darwinian model of consciousness being marketed by Nobelist Prof. Edelman from Rockefeller University. His bold substitution of the classical natural selection criteria of fitness (increased reproductive capacity) for an increased facilitation in synaptic transmission, all based on artificial intelligence simulation and statistical entropy determinations therefrom, leaves un-answered many vital questions, as we will briefly describe below. If he gives great importance to inherited homeostatic controls (which he does) in shaping up consciousness, how does his hyper dynamic Darwinism and re-entrant schemes explain its inheritance, not to mention its subtle workings, especially during the first years of life where exteroceptive input is so limited. The neuronal networks of the visceral brain in the newborn were patterned by executing the instructions coded in the inherited DNA and it is not far-fetched to assume they still play a role in giving stability to Module 1, the other alternative would be to leave this module also subject to the management by his special natural selection, integration and re-entry, all predicated on an assumed background of complexity and 'creative chaos'. It just won't cut, at least for the visceral brain network cluster as evidenced by its synchronous EEG activity (low complexity) and the paucity of its neuronal presence at conscious cortical levels. We have to develop a formulation for the kind of processing the visceral brain does whose non-linear output format evades processing by the 'talking' brain maintaining its participation at non-conscious levels, i.e., non articulable. The linearity of Modules 2,3 output is in our opinion predicated upon the adoption of a binary code format for the "all or none" activity of neurons and this is reflected in the measurement of the statistical variability (entropy) in Edelman's 'functional clusters' aka distributed networks. However, the fact is that many visceral autonomic neurons have graded responses many a times controlled by humoral depolarizers where the threshold value for neuronal depolarization is shifting and changing in polarity. This kind of system will always escape a measurement of the possible pattern of activity the system can adopt at any moment because the number of bits of entropy / per unit does NOT correspond to log (base 2) of (2) as required for binary systems, see discussion below. To the extent that Module 1 is able to have linear outputs, to that same extent the information can be processed by the talking brain followed by consciousness of the event (qualia?). Nevertheless, we highly recommend the Edelman’s work as the first serious attempt to bring together brain scientists and neurophilosophers.
A ‘talking’ silicon brain sans qualia.
History has recorded how often man has incorporated extra-physical elements into his repertoire of strategies to give meaning to the ‘self-evident’ manifestations of physical reality, like the planarity of the earth surface, the geocentric orientation of cosmos or the finitude of the universe. I am not talking about Gods; I am referring to the fuzzy logic of electromagnetic fields, the ether of relativity theory or the counter-intuitive prediction of quantum gravity in string theory. However, these limitations will never stop man from attempting to reduce the low fidelity of his sensory perceptions of the physical environment to formulations he can comprehend and use to make predictions about the future course of events. It is well established the poor resolution of his sensory organs of perceptions and now we have focused on his even more limited combinatorial logic capacity to make sense of those perceptions. Many years ago Piaget struggled with the serious limitations of kinetic equilibrium dynamics to formulate a model of consciousness yet his conclusions are not that far off or counterintuitive as those being proposed by distributive network theory, the new kid in the block. We will try to pinpoint some of its strengths and weaknesses in adopting it to formulate our model of brain function.
The results obtained from the input / output measurements of interconnected network arrays of electronic relays were confusing and un-expected. So was the neuro-histological network arrangement in layers of CA3 area of the hippocampal cortex or the hyperactive complex EEG patterns observed. Whatever chaotic profiles they share in common was captured and harmonized by brain theorists in developing an attractive network theory now adopted to describe many aspects of brain function. We have tried, not too successfully, for the reasons cited above, to restore the relevance of language to thought processes. There are many traps inherent in that approach, we will mention a few.
The most important difference between silicon and neuronal networks has been the non-linear response characteristic of the latter notwithstanding the binary, all or none response of their individual units. Computer operations are essentially reliable linear processes at speeds ca. a million times faster than its neuronal counterpart but degrading catastrophically when a single element fails. This is contrary to the resilient, redundant, variable and interconnected plasticity of neuron assemblies. Furthermore, traditional computers limit their site of operations and memory storage to one or two specific locations as opposed to the layered cortical arrangement. These disparities nearly doomed this experimental approach. Fortunately, a new generation of computers was designed to become more brain-like in their operations, particularly in their distributed parallel processing and the capacity to modify its operations. Unlike the extensive crunching of data generating rigid results, the new generation was more flexible and able to tackle non-linear inputs. The crucial development came when Hopfield’s network design allowed for modification of the output by controlling the strength of the preceding interconnections, the equivalent to the neuro-humoral modification of synaptic connections before the output. This control made possible the attainment of circuit stability from an otherwise random pattern of oscillations. In a Hopfield ‘crossbar’ design, we have a layering of units resembling a cortical arrangement. It is possible to ‘supervise’ the connections by adjusting their strength until a particular configuration is stabilized in a distributed fashion, like the brain does. By changing the adjustments, the same number of network components provided many different configurations each representing a stable memory trace. An operator now can selectively change the strengths of selected locations, i.e., train the network, to achieve desired results. For instance, if an output layer was giving a larger output signal than desired it could be ‘trained’ with specific algorithms by back-propagating the adjustment to the previous intermediate layer > to input layers until a desired output is achieved. Unsupervised networks did not need training and were servo-controlled by its own inner structure; this was reminiscent to the way our Module 1 operates. This led us to consider the possible use of AI to study inter-modular interactivity and to intercalate a language algorithm to influence the network operation in the elaboration of a ‘thought’ image. We later learned that another group had independently developed a way to teach a computer to learn a language; it was called ‘NetTalk.’
In our case, we considered Module 1 (e.g., visceral brain, cingulated cortex) as representing the unsupervised, servo-controlled Hebbian network being initially programmed by evolutionary genetic memory to maintain homeostatic control of physiologic interoceptive variations. Module 2 represented a trainable supervised network modified by environmental exteroceptive, multimodal inputs channeled along parallel and asymmetric pathways. Module 3 represented a balancing act reconciling the immediate environmental requirements with the need to maintain the physical and mental integrity of the individual (‘self’). This first order 'judgment' (awareness) was necessary for the elaboration and execution of coordinated, adaptive, motor responses including verbal communication. The physical proximity of the pre-motor 4s area, Broca’s area and the frontal lobe executive brain area (see Fig. 1) required a sequential input that would have the effect of binding the components of the multimodal imagery that accompanies the motor response. This is when we suspected a language algorithm serving both purposes and started our fishing expedition. We know that Broca’s output articulates the sequential firing of all neurons controlling speech production as we described above. The speech, audible or not, is simultaneous with thought imagery and is maintained by re-entrant and muscle feed back circuits.
As we explained above, such adaptive response includes a conscious (‘talking’ brain), a sub-conscious (hippocampus) and an unconscious (amygdala) components describing diminishing levels of network complexity. The amygdala has an advantage, simpler (operational, not structural) network arrangement and a primitive memory independent (initially) of multimodal inputs, as described above. An experimental design effort is in progress to incorporate elements of NetTalk and Darwin II. The main problem we anticipate with this hybrid is the design of the training algorithm. Our strategy is to substitute the speech-producing effectors (controlled from Broca’s area) with a digital speech synthesizer to match the input parameters (e.g., frequency, intensity, timbre, duration) of the primitive sound (rustling of leaves?), this would be the digital > analog decoder transducer. At the other end the analog > digital conversion is represented by a keyboard arrangement reproducing the cochlear ‘place theory’ generating ‘microphonic potentials’ output to intermediate stages representing inferior colliculus, geniculate body and hippocampus, in that order. We include the CA3 region of the latter because it provides the ‘environmental context’ input during the second cycle as outlined above.
As you can see, computer memory constraints limit the inclusion of Modules 1,3 in the design but we hope to demonstrate how the sound in the original input provides the triggering stimulus to elicit the behavioral response with its alert and affective concomitants, the ‘binding’ factor joining together the multimodal ‘environmental context’ output generated by Module 3. Before all of that, we are working out the solution of the equilibrium states algorithm for all 3 modules under different conditions of ‘supervised training’.
Every strategy for human action supposes a theoretical framework of reference outside of which the conclusions are not validated and become mere speculations. In brain science the theoretical framework is, of course, the result of a scientific methodology based on the direct observation of nature (sometimes aided by instruments) or of nature as simulated in the laboratory. In either case, we strain to inferentially accommodate the observations within the constraints of logical formulations, be they logical sentences or symbolic equations. This is so because that way they can be subjected to combinations, permutations, contrasts, comparisons, relations, etc. or any other logical operation that may result in practical adaptive forecasts of future events.
Throughout recorded history two events have remained constant, the use of language signs (sounds, words, music, body movements, etc.) for communicating our experiences 'logically' or appeal to extra-physical explanations when they elude being fitted into conceptual, logical strait jackets. Either way we quench the thirst for explaining our world.
Rather than fight the obvious we have always approached reality with three working hypothesis: 1) man’s unconscious priority is his biologic and mental survival and that of his species. 2) this goal is accomplished with an eco-logic and an anthropo-logic elaborated by the Turing component of his brain and a 3) theo-logic to accommodate the extra-physical presence. The first hypothesis is self evident under behavioral optics, the third hypothesis is well supported under historical optics and the second one we try still labor to use as a bridge between the other two in a comprehensive, coherent, consistent and credible fashion, not necessarily an absolute account of the Kantian ontological reality in itself but a working reality. To conclude that the extra physical reality will eventually yield to technology and be reduced exclusively to formulations or symbols is an act of faith. It is irrelevant to put labels on the strategy we choose to labor in the achievement of social conviviality.
*Fig. 1 taken from “Visceral Brain, Language and Thought.” Noesis 2001
** Review article Amazon.com 200
End Chapter 8