The Neurodynamics of Causation.
*The research on the cortical ‘attractor’ model for olfactory inputs by Dr. W. Freeman has been a seminal source of valuable information in the unraveling of the super complex dynamics of brain function which to this date remains unsolved, especially when we try to modify and adapt/expand the olfactory model to include and functionally integrate ALL perceptual (sense-phenomenal) and conceptual (symbolic or sentential logic) functions as discussed in previous publications. We have really tried to remain close to measured or mathematically-derived data, from EEG, MEG, fMRI instrumental data to self-evident experiences like perceptual and/or conceptual qualia. Perhaps the most difficult step will always be, not so much the explanation of how perceptual/conceptual inputs are amplified by brain background signals to form new and update old cortical attractor basins by quantum wave resonant/amplitude couplings or how they parade before our mind as a flow of consciousness, but how do we isolate one of several such cortical attractors (future outcomes) as the exclusive conscious choice based on its probability of adaptive success in response to environmental contingencies. But we have no doubt that, just like we consciously select the most effective motoneuronal networks -among so many available- to safely score points at Olympic gymnastics, we can also freely choose to isolate the equivalent cortical attractor neuronal networks when adapting to those environmental contingencies. But we ask, how does the non-physical mind is able to be causally efficient on the physical brain during the conscious exercise of free will, what we have opted to call ‘proximate causation’ for lack of a better characterization? We may never know how many elements, if any, functionallt mediate or anatomically bridge the invisible transition from the conscious choice to the motoneuronal effector network(s) or the relevant mental state. This explanation, within the context of brain dynamics, will have to be metaphysically conceptualized as a problem of complex indeterministic causation or else surrender your analytic effort to the accommodative label of ‘emergent’ psychophysical behavior or a ‘functional’ approach. We outline below some of the gray areas resisting being framed into conceptual formulations and revisit the ‘conditional probability’ tool within the context of neurodynamics.
Probabilistic or indeterministic causation remains refractory to being successfully reduced to operational algorithms because of so many complex gray areas. Before throwing in the towel, we should explore causation in itself while zeroing on functionalist aspects according to which we should restrict our view of ‘mental states’ as those defined by their causes and effects. This focused view excludes, for the moment, any consideration of intrinsic properties of such states but rather concentrates on the mental state in relation to immanent or transcendental perceptual and/or conceptual inputs, other inner states or behavior. This narrowing has the advantage of conceiving mental states as being conceptually shared and realizable by different systems such as computers and even other biological species..
Every long journey starts with one small first step, what do we understand by ‘causation’ as it applies to brain dynamics? In general terms we may think of causation as an act, intentional or not, that brings about an effect, be it an object (e.g., work of art), an event (e.g., punishable behavior) or a mental state (perceptual/conceptual qualia like sounds or sadness, respectively). Functionally speaking, for each and every consequent act of causation there has to be an origination in the form of an agent (person or object), an event (e.g., a force causing change in motion) or an antecedent mental state. We will use the term ‘sine qua non’ to imply that the agency, event or mental state initiating the relevant end result, must be both necessary and sufficient to have caused such consequent response. We will briefly discuss below the special case of ‘propositional attitudes’ *(Fodor). Having now framed the scope of causation we are most interested in, let us illustrate with some further focusing examples to follow.
We have inherited motor reflex arcs (with varying degrees of complexity) to preserve the biological integrity of the human species, from the simple monosynaptic, segmental withdrawal flexion to the complex, polysynaptic, suprasegmental walking reflex. It always involves a receptor ‘organ’, an effector (muscle, gland) and sometimes intermediate associate neuronal connections. These basic varying connectivities, by and large, are genetically hard-wired and cannot be consciously accessed at their unconscious level loci to modify but we may sometimes consciously inhibit their instantiation (against self interest) as when terrorists explode themselves in crowds, religious fanatics walk on nails or activists set themselves to fire as we witness during pathological, heroic or altruistic acts. Consider the following specific example to help us further focus on brain neurodynamics.
We may have noticed or heard that an unconscious anaesthesized patient in the operating room (OR) will not reflexly respond to remove the nuisance produced by slight mechanical touch stimulation on his cheeks caused by the neurological hammer, a bristle, a pin prick or a walking flea. The observed behavior is different when the same events are repeated while he sleeps in the recovery room or when fully awake during vigil, now he responds with a slap to remove the nuisance stimulus. What does that tells us? Unlike the OR where the deep stage of anesthesia inhibited all motor responses, in the recovery room we are dealing with desinhibited segmental reflexes possibly involving a few spinal segments controlling nearby superficial, touch receptor areas on the cheek. Each small area with touch receptors (dermatome) is wired-up to specific motor neuron pools activating one among varying neuronàmuscle combinations covering the moving receptor surface area. Another significant difference is that the sleeping unconscious patient may not reflexly respond by self slapping his cheek to remove the nuisance unless the position of the nuisance (e.g., flea) changes whereupon the hand precisely aims at the new location and activates the appropriate motoneurons to slap the flea down. This unconscious isolation of one of many probable motoneuronal responses (depending on surface location of stimulus) illustrates the diversity and precise network configuration of inherited responses available. This interpretation is all based on classic neurophysiology after the British school of *Sherrington, et al.
Lets introduce now another variation, let’s wake the patient up. Suppose the same aching patient receives same nuisance flee stimulus while he is awake recuperating from the surgical spinal decompression procedure to remove the neck pain from a pinched (as it exits from the cervical vertebra) upper cervical peripheral nerve. Why isn’t the consciously inaccessible, gene-controlled reflex arc is not unconsciously triggered into action this time, like before when he was asleep? Because now, if it did, a reflex slap on the flee area of his face would likely unintentionally jerk his neck and undo the surgical relief previously provided. But instead he consciously willed into action a different otherwise inaccessible & unconscious segmental reflex that produced instead a milder contraction of his facial musculature. Had he responded and jerked his injured neck, an amygdaloidal, gene-controlled, neck avoidance reflex would have aggravated his condition. Fortunately, while the amygdala is temporarily inhibited from reflex responding to the cheek stimulus, the cheek stimulus also traveled to the meme-controlled memory data base in the hippocampus for a ‘context analysis’ of the situation. This resulted in a subconscious level comparison analysis of the biological risks involved in the various probable responses available in the cortical attractor basins connected to the hippocampus. Consciously isolating and activating the relevant cortical attractor controlling local movements of facial musculature (e.g., buccinators muscles) was a more conservative choice of response than the unconscious reflex release of a damaging slap controlled by the amygdale, now inhibited from acting while waiting for the result of the hippocampus context analysis. The meme-controlled memories or conceptual analysis would oppose the slap & neck displacement to ‘preserve’ biological *integrity of the patient and a more conservative cortical attractor alternative was instantiated. As we have detailed in other publications notice how gene-controlled reflex neural networks at unconscious level are transiently inhibited when *stimulated by potentially life threatening stimuli (See Le Doux avoidance reflex) pending a context hippocampus evaluation at subconscious levels. Based on meme-controlled hippocampus memories, the gene-controlled amygdaloid memories are released from inhibition and a ‘fight or flight Cannon response’ ensues to protect the species. If the contex analysis reveals no such danger (e.g., a sudden poisonous snake rattle sound was coming from a cage in a nearby zoo). In this case the context analysis posed no biological threat to the actor and no cortical activation of an attractor solution was necessary.
It is important to notice the obvious general similarities in the cortical attractor adaptive response to perceptual/conceptual inputs when it includes in addition other extrabiological, psycho-socially acquired aspects of existential reality. This time the biological gene-coded memories in the amygdaloid-limbic system neuronal networks are synaptically integrated to their appropriate psychosocial, meme-coded memories in the hippocampus network complex by variable strengths Hebbian synapses. The genetic aspects of musculo-skeletal control by motor-neuron probable responses can accept limited modifications/updating (through physical training) of the future motor outcomes pool). These infrequent alterations are more primitive in the sense that, while defaulted by biological preservation imperatives (e.g., visceral neuro-humoral equilibrium), modifications do not normally require access to conceptual memory pools (logic/language processors) for subconscious context analysis except as noted. Functional musculo-skeletal updates by exercise training are consciously processed and decay when not incorporated in a lasting memory data base. There is a notable exception when a subconscious context analysis report, that would have resulted in the eventual conscious activation of an appropriate cortical attractor, is consciously rejected in favor of a conscious will to act against self-interest, contra-natura, as in the cases noted above. We have seen how unconscious, gene-controlled biological preservation and meme-controlled psychosocial equilibrium fundamentally differ
in their inputs and context analysis before the intentional isolation and actualization of their ‘future outcomes’ neural networks. Like in computers, we have access denied to core company-inherited programming while being able to access, modify, isolate and execute those we personally programmed and choose to execute at will. Notice though that in both cases we CAN control by proximate causation, e.g., turning the PC off or changing programs! But notice also that, unlike the computer case analogy, we still ignore the primitive neurophysiological mechanisms underlying the conscious choice of BOTH the appropriate motor reflex in a complex postural feat AND the appropriate cortical attractor from a pool of probable alternatives. At the reflex level of musculo-skeletal control the maintenance of bipedestation and an erect posture when opposed by disrupting gravitational forces (as in gymnastic movements) trumps any other body posture during exercise in the normal person, except as noted above.
We can arguably compare the segmental and extrasegmental moto-neuronal pools as the neuro-anatomical equivalent of the cortical attractor networks and label them both as representing future outcomes available to choose from, at conscious or subconscious levels as illustrated. We can even extend the analogy to suggest that the motoneuronal menu has varying probabilities of selection depending on the immediately preceding body posture just like the probability of selection, isolation and activation from an attractor menu is predicated on appropriate antecedent perceptual/conceptual input as discussed. What remains for both sub models to explain is just how our mind zeroes on the appropriate physical brain network to activate it into actuality? In our opinion this is clearly a problem of unfathomable probabilistic causation where complex ontological and epistemological issues control our explanations as we will elaborate on below before settling on a *modified Fodor’s ‘propositional attitude’ model as discussed elsewhere. We proceed to lay the foundations for that search. It seems appropriate to keep in mind that reality ‘in se’ dispenses of any epistemologically imposed constraints, whether *temporal, linear, symmetrical, or causal ‘corrections’, as we discussed in two previous writings about how our species inherited the ability to incorporate an epistemological temporal clock to harmonize the primitive atemporal, asymmetric, non-linear, perceptual/conceptual environmental inputs with our language-controlled linear and sequential way of processing such data. Fundamental physics does not need any such ‘corrections’ considerations either, as Russell and the modern ‘eliminativists’ have argued. So causation is a mind-created concept to extract meaning from a seemingly indeterministic or chaotic reality. It is useful in that it allows the human species to reductively analyze such reality within the perceptual and conceptual limitation genetically imposed. We aim at following a reductive, conceptual, analytical path between the extremes of Russelian eliminativism and primitivism as long as there is lacking scientific criteria to distinguish between a conscious ‘cause’ of a mental state or thought and the latter’s alleged intrinsic condition of self-generating the state sans causation (emergence). We do admit that conceptual modeling often requires the selection of that self-serving and convenient condition to play the role of ‘cause’ regardless of its truth value. In this respect, of considerable importance to neuropsychiatrists, there exist verifiable fMRI data documenting the state of limbic-based emotional euphoria. With the ancient Greeks, we believe that achieving a maximal mental state of happiness compatible with concurrent biological integrity, if possible, is the ultimate goal of existence. Accordingly we speculate that the cortical attractor to be ‘consciously chosen’ from the various alternatives available plays the role of being itself causally efficient in instantiating the happy mental state or vice versa, when a given mental state, controlled by neuro-hormonal current states, acts to select, isolate and instantiate the compatible cortical attractor, in a reciprocal dynamic manner. Whatever the final answer may eventually be we must continue to scrutinize neurodynamic causality. If we analogize the conscious choice (free will) with the instrumental collapse of the wave function during a quantum mechanical measurement we need to carefully examine also the probabilistic causation variant.
Unfortunately, in our examination to follow we will have to introduce less than universally accepted terminology and concepts in this leading edge, ambitious philosophical analysis as a mere guide to travel the choppy waters of brain causation. At first sight it may seem easy to explain causality when simply describing a tennis racquet hitting the ball towards the net as a simple event category involving two entities whose roles are clearly to cause and to respond, respectively (C causes the effect E). But there is more in the causal relata than a willing agent and a visible result when explaining the neurodynamics of probability causation (e.g., perceptual C rather than conceptual C* input causes mental state E rather than motor behavioral effect E*). Probability causation should include, besides the perceptual or conceptual input, a categorical characterization of the consequent object, event or mental state that results, singularly or together, and their respective roles when they become a behavioral or mental state reality. To complicate matters further, the category may refer to either spatiotemporal immanent results being instantiated (e.g., ongoing, online act of painting) as opposed to transcendental result (e.g., the abstract, non spatio-temporal, off line, cognitive fact that such painting exists in my memory). A subsidiary aspect of the category is its granularity, whether the result is individuated (countable, coarse granularity) or generalized (invisible structure of Bohr’s atom of fine granularity). This is important in establishing the nature (linear, cyclical, transitive, etc.) of the causal relationship, if any. A related aspect in the granularity analysis of the chosen cortical attractor result/effect is the reliability or truth value assignation. The result may be connotational (e.g., a contingent proposition, suggested or implied meanings/attributions) or denotational (e.g., predicate concept, exact literal meaning). What this means is that if the agent chooses a general abstract solution of fine granularity as his result, the causal relationship is termed intensional or connotational (e.g., an explanation) but if the choice is more specific, of coarse granularity, then the relationship is termed extensional (e.g., a description) or denotational. Since neurodynamics implies a conscious choice among alternative cortical attractors based on their probability of implementation we can say that during the flow of consciousness, before the conscious choice, the causal relation result is intensional/connotational and as soon as an isolation and a conscious choice is made the result/solution may be anything depending on the level of organization being considered. I may be thinking of choosing a specific missile weapon of coarse granularity, descriptive and denotational OR choosing the best equation of motion describing its trajectory in general terms, of fine granularity, explanatory and connotational, as the causal relationship may be. Thus, when a presidential candidate S consciously chooses the alternative (among several probable outcomes available) of an unconditional pull out of military personal from Iraq predicated on his ‘belief’ that P the local government ‘may’ control terrorist activity [an example of Fodor’s “propositional attitude” (S believes that P)], the causal relation result is intensional and connotative as opposed to the other candidate S’ choosing to keep troops there for 2 more years (predicated on his factual ‘knowledge’ that the Alliance commitment to immediate deployment of 10,000 troops ‘is’ viable), is an example of a different propositional attitude (S’ knows that P) where the causal relation is extensional and denotational. S explains, S’ describes the probable future outcome. Notice that the granularity of the attractor’s perceptual/conceptual content itself controls the causal relation between the agent (presidential candidate) and his conscious choice of a solution from the available probable future outcomes. To the best of our knowledge none of the available monitors of brain activity, including fMRI, can experimentally distinguish between connotational and denotational attractor content (and indirectly describe/explain the causal relation), a psychology experiment is sorely needed. Needless to say that either intensional or extensional, connotational or denotational a choice may constitute a case of perceptual/conceptual underdetermination or overextension. In the search for meanings to be extracted from an instrumental measurement of an environmental object/event (ontological object), the observation may be forced to loosely fit into a mathematical formulation (epistemological abstraction) that may either underdetermine or overextend the fine-grained granular representation of reality ‘in se’. Existential reality is consequently an operational epistemontological hybrid in an undeterministic but probabilistic world. In another writing we may express our views on the extreme cases where an abstract, conceptual representation of invisible ‘reality’, i.e., string/’brane’ theory, has aspiration on being accepted as an unique case of ontological reality!
There is a related complication where the fineness of the granularity may never be ascertained because of the lability of interacting cortical attractors in the basin whose individualized neuronal network composition may vary in perceptual and/or conceptual content, making it almost impossible to instrumentally determine the sequence or transitivity of synaptic events in the causal chain. Considering the speed with which these events happen, may as well ignore the transitivity in intermediate causality, linear or recursive. Better to bite the bullet and settle for coarse spatiotemporal ontological individuation when representing events.
As promised above we now proceed to briefly examine some relevant points on the probabilistic causation variant as it pertains to brain neurodynamics. It is possible that the quantum neurodynamics control of indeterministic causation may sound counterintuitive because causality is usually associated with determinism (even in the natural sciences). As we said above, this is a special case of indeterminism because it carries the implied presumption that a conscious choice can cause the preceding indeterministic process to become determined by the isolation and execution of a specified event based on its probability of being *selected among other existing alternatives. As we elaborated in another writing, when an object, event or mental state, acting as a perceptual or conceptual input, raises the probability of their effects as expressed in the conscious isolation and selection of an appropriate cortical attractor among various other future outcome alternatives (the equivalent of an instrumental measurement causing the collapse of a wave-function in quantum theory), then we can formally represent the situation in a ‘conditional probability’ format. If we suspect that various perceptual or conceptual ‘factors’ stand for potential causal agents (causal relata) we can formulate their potential status as causal agents on a probability basis as: P(B | A) meaning that the conditional probability of cortical attractor B being empirically isolated and instantiated (based on observed behavior) given the particular perceptual/conceptual input A acting as causal agents. Formally expressed as a probability function we get the probability ratio: P(B | A) = P(A & B)/P(A). This *approach, as we have detailed elsewhere, has the advantage that it does not have any dependence on temporal, causal or symmetry constraints, in harmony with reality ‘in se’. Ideally the indeterministic probability characterizing brain function cannot be conditioned on a tautology relation without denying its dynamic nature. However, probabilistic causation theorists have incorporated the counterfactual notation and techniques to remove spurious or otherwise irrelevant causal *influences in an attempt to reach the tautology ideal in the analysis. See Markoff Condition. A formal elaboration is beyond the scope of this brief analysis.
Probabilistic approaches have their own built-in limitations as is the case for ‘pre-emption’. Imagine two neuronal pathways with different synaptic weights where A, the least probable to cause an effect (slower conduction) initiates the action that randomly activates a component of the faster second neuronal pool B that causes an effector response pre-empting A from arriving at the target. Which should we attribute as having raised the probability of a target response? Which of the intermediate neurons in the train of causation was randomly activated? This problem is particularly relevant due to the very nature of attractor networks, their lability of connections is such that an invariable pattern of succession/progression along intermediate synaptic components cannot be guaranteed unless indicated by appropriate brain monitors, e.g., fMRI, as they happen, barring the occurrence of causally irrelevant pseudo connectivities that are concurrently activated. A related complication arises when a perceptual/conceptual causal input may not be followed by its expected probable effect, e.g., spurious (random or with regularity) inhibition. Causation is usually associated with asymmetry, i.e., cause C à effect E is not reversible. As mentioned before, we have postulated two instances of possible reversible causation: retrocausation where recursive cycling in neuronal pathways brings about update modifications originating in cortical attractors into a subject’s thinking instead of the other way around. A second interesting reciprocal symmetric causation is illustrated when a neuro-hormonal based euphoric emotional mental state may cause an induction bias in favor of a particular cortical attractor as an adaptive solution to a contingency being experienced and viceversa when a consideration of that particular future outcome produces happiness. An important element of our modified cortical attractor model is predicated on the reciprocal causation between the continuous perceptual/conceptual updating inputs into the attractor basin and the associated mental state it generates.
SUMMARY AND CONCLUSIONS.
In this occasion we have revisited for the third consecutive time the complex issue of indeterministic causation within the context of brain neurodynamics, from the configuration of attractor basins in the human cortex from sense-phenomenal, perceptual and memory based conceptual, never ending updating inputs (that bring about fleeting moments of consciousness flow evidencing our stock of probable future solutions) to possible environmental contingencies as they may arise. We dedicated the most space to revise the relevant aspects of the metaphysics of causation to provide the guidelines for a systematic approach in its analysis when we opt to consciously isolate and choose, from many probable alternatives, the attractor best suited for a contingent solution to environmental challenges. After such systematic perusal of issues to be encountered and its possible formulation for analytic reductions we come to the troubling realization that, because indeterministic causation in brain neurodynamics requires examining the intermediate labyrinthic neuronal network pathways between the conscious causal agent and the resulting effect in the isolation and instantiation of the chosen attractor, we tend to single out self serving conditions that best suit our biases without the benefit of a determination of their respective truth values or an experimental confirmation using the appropriate brain instrumental monitoring or observed behavior. Another disappointing conclusion pertains to the coarse granulation understanding of brain function probabilistic causality imposed by our analytical tools, instrumentally and conceptually unable to identify extensional/denotational singular causality, leaving us instead with the uncertainties of general or intensional/connotational causation. The future of brain neurodynamic causation is predicated on the success of the causal relata being formulated in propositional format for ease of computer simulations. We are not aware of any serious work beyond Fodor’s ‘Propositional *Attitude’ model formulated around his ‘Language of Thought’ hypothesis. This would allow us to examine our tentative speculation on the reciprocal symmetric causation between neuro-humoral instantiated emotional states and the cortical attractor about to be chosen, among other things.
Dr. Angell O. de la Sierra In Deltona, Florida Summer 2008