V

WHAT DO WE KNOW ? − Cognitive neurophysiology

Doctor Franz Joseph Gall, at the beginning of the nineteenth century, had already understood everything.

Here is the area of comparison, there music, time, color, order, weight, dimension...

Here is calculus, if one has this area more developed he is more clever in computing and the skull has a bump here: the “bump of mathematics”.

He had already explained us everything.

And please do not say that these are only cognitive capacities, nothing to do with the soul, because there we also have seriousness, and hope, ideality, spirituality, here love, parental love, even conjugal fidelity − if one does not have the bump...

Soul, uncertain word. One of those words like life, intelligence, memory (love?), that change their meaning and color depending on the moments and the context, that have a different meaning − a thousand different meanings − for each of us.

Difficult to grasp. Now, looking at the brain, it appears equally rich of processes, modes, approaches, representations and interpretations. Capable of innumerable multiplicities, simultaneous and contrasting, and a thousand levels of analysis and synthesis, and meta-analysis and re-synthesis, and re-discussion.

A science that dares study such a complex, intricate and multifarious system cannot recoil in front of the commotion of complexity and infinity.

The brain takes care precisely of this, facing complexity ad infinity: it transforms each information that reaches it, and reads it simultaneously in a thousand different ways. It decomposes and reunifies elements and relationships, thereby building not only representations but criteria and interpretations.

May we define the capacity of transforming sensory data into a representation of elements and relations as a process of “abstraction”? Immanuel Kant would possibly recognize in it an unexpected materiality of his beloved process of transcendental knowledge: the application of a “filter” to sensory experience. But then he should admit that at least some of the “categories” that constitute the transcendental scheme − that should mediate between the sensory perception of the phenomenon and its intellectual conceptualization − are interposed well before even a true sensory perception is reached: categories (more or less rigorous) such as Unity, Plurality, Negation, Limitation, Cause, Community, Necessity, Analogies, are immanent to the neuronal schemes themselves that re-elaborate sensory signals.

It might be more precise, then, to talk of translation. If there is an intellect that interprets sensory data, these have already been translated before they arrive here, according to a series of re-elaboration schemes not at all elementary. These same re-elaboration schemes are real and material (they can be recognized in neuronal circuitries) and as such they are object of scientific investigation. An investigation that has undergone an impressive acceleration in the last twenty years.

As already mentioned, the main difference between man and the other animals is the extraordinary development of multimodal associative areas, regions in the cerebral cortex that take care of combining readings performed by other areas and “interpreting” them, by integrating information that comes from different sensory modalities and have different relevance: visual, auditory, visceral, motor, emotional information... The regions where pieces of information of different nature converge extract properties that are not present or detectable in the single sources and modalities of information: it is a novel reinterpretation, multiple and multifaceted. Most relevant, it is a different reinterpretation in each of the “multimodal” regions of the cerebral cortex, thereby generating a multiple reading that adds emphasis, thickness and depth to raw and flat information, as the standpoints and complexity levels multiply.

The fundamental result of this process is that each sensory datum − object, phenomenon, situation − is translated into the activation of a certain number of neurons in the cortex; on the one hand, this scheme of activation practically is the cerebral equivalent of that sensory datum; on the other hand each neuron − or group or neurons − that participates to such scheme represents a feature of the datum itself, and will be activated also by other sensory data that present that same feature − and will also participate to the activation schemes that correspond to such other data.

In a sense, it is like translating a drawing into its description: there, a rounded pinky shape, with two blotches of shade surmounted by black dashes, a central prominence, an extended horizontal reddish mark that moves, changes its shape and sometimes lets through a number of small whitish blocks or a central darker region, and the whole surrounded by a scrawl of un uncertain dark color, and slightly to the right... Those who play with the computer know that it is possible to draw on the screen by using, in place of a virtual brush, instruments that generate squares, triangles, circles, polygonal or curled shapes, filled or empty, colored as one wishes, possibly with shades. It is a way of drawing that saves room on the computer disk, in comparison to creating a file where an image is represented by a value of red, green and blue intensity for each of its points (pixels). It is a way of drawing that lets you modify elements and relations with great facility, without erasing and redrawing, simply by changing the geometrical features, the color, the visibility properties of the graphic objects that constitute the drawing.

The most relevant aspect of this translation of each experience into a scheme of neuronal activity possibly lies in the fact that the activity of certain neurons or groups of neurons represents the presence of specific relations in the sensory experience. An example: all somato-sensory information − tactile and skin sensibility, and “proprio-ceptive” information, that informs on the position of muscles and joints − arrives to the anterior part of the parietal cortex, ordered in a topological way, so that they draw a virtual map of the whole body on this region of the cerebral cortex; from the occipital cortex, posterior, visual information arrives and is mostly interpreted in terms of spatial relations, as it proceeds forward moving away from the primary visual area toward the posterior part of the parietal lobe; the central region of the parietal lobe, interposed between these two regions, simultaneously elaborates and compares information on one’s own body and on the external space. Several groups of neurons in this region are therefore in charge of detecting relations of proximity, distance, repetitiveness, order: each group will be activated every time one of this kind of relationships is detected in space, and its activation will precisely represent the occurrence of the relationship, the experience of the relationship, the relationship itself (the concept?). But it is clear that, as the complexity of elaboration grows, other specific neurons will elaborate, rather than simple spatial relationships, the reciprocal rapports among such relationships, until the objects and events of experience are framed in a complicated network of spatial relationships, that essentially constitutes the idea itself of “space”. Yes, the conscience of spatial relationships precisely arises from the activity of these groups of neurons: a subject that suffers a lesion in this area of the cortex, on a side of the brain, loses any interest for spatial rapports in the opposite half of the space (information that arrive from the left half of the body are processed by the right half of the brain, and vice versa); if the lesion is on the right he fully loses the conscience of the left half of the world: he hits obstacles and furniture that are located on the left, and if you ask him to reproduce the drawing of a flower he will reproduce the right part of the flower only; still he can see, very well, because if you propose the drawing of two flowers he does reproduce both flowers, but only the right part of each...

A scheme of neuronal activity for each object, a scheme for each relationship; in a sense, a specific scheme − those neurons activated that way − for each concept. This might disappoint, but the right word is exactly this one. In regions close to those that elaborate spatial relationships there are groups of neurons that recognize spatial sequences and, in particular, forms of order, and ordered series: each element close to the next, and each greater than the next, or smaller, or darker, or nearer... It might astonish, but these neurons and circuitries are also “used” to elaborate information that comes from other regions of the cortex and has nothing to do with spatial relationships and sensory information: their activity is indispensable to count, enumerate, recognize hierarchies (of position, or dimension or luminosity, but also of age, importance, emotional relevance): every time a mental operation involves some form of counting, from numeration to more sophisticated mathematical computation, the neurons of that region are the ones that show intense activity.

In general, although more roughly, the modalities of information processing are essentially similar in less developed animals: the sophistication, depth, versatility and precision of most processes of recognition of features and relationships are more rudimentary. The crucial difference, however, does not lie in that, but rather in the intensity and richness of multimodal elaborations, that is in the capacity of relating in a thousand ways information of different kinds − visual, auditory, tactile, kinetic, proprioceptive, visceral, emotional − toward the creation of schemes of neuronal activation that correspond to that specific combination of sensory and relational (cognitive, in a word) aspects that synthetically characterizes a concept, that abstract entity that lets us classify dogs as different from cats although all have four legs, hair and a tail − and a Chihuahua is more similar to a cat than to a firedog or a Neapolitan Mastiff; that abstract entity that lets us classify a photograph as different from a painting, or a dream as different from a desire.

The impressive development of associative multimodal regions in human brain cortex creates the possibility of an almost infinite multiplication of concepts that can be represented in their complexity, in their finest differences and their most subtle, intricate and delicate relationships. This enormous development is accompanied by another big evolutionary jump, the paramount progress in the possibility of modulating sounds, thanks to a vocal apparatus that is much more sophisticated than in any other animal, and to the large expansion of the portions of cortex that process the emission and the perception (and elaboration) of sounds. It is not easy to tell which of these two developments has preceded and possibly favored the other one: presumably, each of the two has made the other one possible, the richness of the sensory and cognitive material to be elaborated and the complexity of circuits and modes of elaboration have grown in parallel, and not all of a sudden. But it should be clear that we are talking of elaboration of sound perception, because this is what is much more sophisticated in man, not the possibility of perceiving sounds: a monkey, or even a mouse, does not hear less or worse than we do, it has not any greater difficulty in distinguishing two different sounds, but it elaborates such perception in a much less complex and refined way.

The extraordinary capacity of relational and interpreted classification, together with the availability of an infinite repertory of sounds (phonemes) that can be combined into precise sequences, creates the possibility of associating a sequence of phonemes (a word) to each pattern of neuronal activation − to each sensory concept − and to each relationship between such concepts; these combination of phonemes can be reciprocally linked by interactions − phonetic, morphological, grammatical and syntactical bounds between and among words − that can reproduce any possible relationship among objects, events and relationships themselves: each possible, observed or merely imagined relationship, or even absurd or impossible ones.

The production of such a system of abstract relationships among elements, that are arbitrarily associated to objects, relationships, acts and events, constitutes a symbolic system. This is the fundamental feature of neuronal elaboration in man, that is made possible by the motor and phonetic versatility and by the substantial development of the multimodal cortex: the symbolic manipulation of reality becomes possible.

One particular region plays a fundamental role in the production of the most versatile and powerful symbolic system that is generated by the human brain, language. Those who have diligently studied the paragraphs about neurophysiology in pills, above, will have no doubts on where this area must be located: in the lateral portion of the cortex, somehow in between the areas of visual, auditory, somesthesic elaboration, close to the limbic system as well, and possibly sufficiently anterior to be able to consider information about motor schemes as well. This way, here it is possible to collect the information about shape, texture, sound, possible practical use, emotional and affective value of an object, and to associate to it an ensemble of sounds − a word − that represents it; a word, both as we hear it with our ear and how we pronounce it by means of a complex modulation of movements of the larynx and tongue. This area is called Wernicke’s area, and it is fundamental for language, not so much for phonation (emission of sounds and words), but rather for the capacity of correctly choosing words, of correctly putting them in relation to express a concept. A lesion in this area does not impair the capacity of repeating a sentence perfectly, but prevents the formation of an original sentence without confounding sounds in the formation of words and without making errors in choosing the appropriate words. In is intriguing how small lesions in the temporal portion of this area (the superior portion of the temporal lobe) interfere with the ability of fetching and using specific groups of words, such as for example those that refer to actions, or working instruments, or geographical names or people’s names. This is one of the most disheartening aspects for those who refuse a “neuronal” view of cognitive functions, because it confirms in a disappointing way how the organized and classified memory is precisely based on fixation of learnt responses into precise neuronal schemes, well localized in a certain region of the nervous mass of the brain.

Presumably, the phonetic versatility in particular makes the symbolic leap possible, because it brings about an extremely efficient − precisely, symbolic − translation. And this conquest in turn pushes and sustains the further development of cortical areas that are suitable for symbolic elaboration, from the first prototypes of homo to homo sapiens sapiens (who knows, might the next be, perhaps, homo modestus...?).

All this does not imply that without speech it is not possible to develop symbolic capacity: the human brain is so rich of such possibilities that it can use any modality of symbolic coding − from mimic and skeletal-motor gestures to drawing and writing − to that aim. Simply, without verbal coding, and without this mutual enhancement by symbolic elaborative capability and speech-related motor-sensory proficiency, the evolutionary development of human cortex would not have been so impressive and revolutionary.

The generation of the capability of symbolic representation through the precise working modalities of specific neuronal systems not only is fascinating − at least for those who suffer from that deformation of intellectual curiosity that make them sensitive to the appeal of understanding how and why. In addition, it also evokes a subtle form of wonder, and the clear feeling that something big is going on here. Something that does not allow us to go on reasoning in terms of connections, functional relations, neuronal networks and strictly mechanistic processes. When a system like the brain learns to manipulate symbols instead of information, the possibility arises of not only representing reality, but also of interpreting it, of imagining it different, of finding answers to questions such as “and what if now it happened that..?”, of inventing, this way, the instruments and the strategies to change it.

A new dimension is born this way, in human brain; the world of ideas is born, that Plato imagined hidden in a cavern, as the site of true reality, invisible to man but through the shadows of the perceptible appearances of the material world. It is an intriguing overturning, this one. It suggests that between our senses and the world of ideas there are indeed complex and relevant mediations, but that such mediations are in our own brain, have biological and physiological bases and explanations, and can be studied and clarified in ever greater detail. It remains true that we cannot reach any certainty about reality out there, our senses may deceive us. The world of ideas, in which our intellect rambles, would then be a construction of our own brain that does not reflect at all the deepest and most precise essence of external reality, as Plato used to like believing. Sure it is possible. A margin is there of unknowable, nobody could deny this, a philosophical domain, or better a mystical domain, where the flight can only be blind, and perhaps because of this even more exciting: the world and life can be imagined different from how we see and live them. He who likes this has the full right to play this way. He has the right to fly away, let’s hope he will find his window open when he decides to fly back, and will not be forced to commute for ever, like Peter Pan, to and fro Neverland...

In the complex schemes of neuronal activity that represent aspects of the external reality (objects and relationships between objects) various types of sub-schemes can be identified. On the one hand, a specific pattern of activation of certain neurons will represent persistent aspects of the object (or relationship), essential aspects: similar objects (relationships) will also display such aspects, which can be generalized until it becomes possible to classify and recognize as “similar” any objects (relationships) that also display such traits, even though they have never been encountered before, just because they are capable of producing these same patterns of neuronal activation. It is like saying they are the same thing (please, pick a name to indicate them all). On the other hand there will be sub-schemes that represent characteristics of the objects (relationships) that may or may not be present, qualifying sub-schemes; these aspects may also be shared by other objects or relationships, though they are not essentially similar. They are accidental features (attributes) that may or may not be there and apply to the object. A further kind of sub-schemes, finally, will represent relationships with the context or with other objects /relationships: spatial relationships, or similitude /difference, relationships of collective behavior, of association or sequence in past experience, and therefore of possible causality...

Actually, this kind of semantic classification of neuronal sub-schemes that constitute our experience may well remain implicit, unnoticed, unstated. May not be verbalized. It may still work perfectly as an instrument for an implicit reading and interpretation of reality, through rapid adaptation to new experiences, by means of the recognition of previously acquired schemes and their manipulation to adapt them to the new situation, and through anticipatory behavior based on the predictions offered by the application of known schemes to the new situation. An implicit intelligence arises from this, a non-verbal intelligence, that permits solving operational problems, even pretty complex ones. Often it is not necessary to explicitly understand, and even less is necessary to be able to explain.

The animal halts here.

Conversely, when symbolic manipulation flanks associative mechanisms, comparison and generalization, then the situation drastically changes. The coordinated and systematic association of a sign to each scheme or sub-scheme of neuronal activity automatically implies the generation of an intricate and sophisticated system of relationships among the signs themselves, of a structure that reproduces the formal features of the system of relationships that exists among neuronal activation schemes. A true symbolic system, in which each sign becomes a symbol, i.e. is given not only a meaning but also a reticule of precise and inevitable relationships with the other symbols.

Two general consequences derive from this, on the symbolic system as a whole: a morphological one and a structural one.

The morphological consequence: inevitably, entire families of symbols will be generated by variations on the theme, by applying standard transformation rules to each symbol. In language, for example, this translates into the possibility of adding prefixes and endings, repetitions and accents to indicate variations of gender (masculine/feminine, in many languages), number (singular/plural), temporal location or modes of an event or action (mode and tense of a verb). A simple principle of economy will inevitably lead to this kind of evolution of the symbolic system, since each datum of experience can present itself with variable characteristics and will therefore generate a scheme of neuronal activation that can exhibit multiple alternative sub-schemes: these variable characteristics are easily translated into morphological modifications of one and the same symbol; the same variable characteristics, recognized in other data of experience, is easily translated applying to the corresponding symbols the same (or similar) morphological modifications... Full families of new symbols can be generated this way by analogy.

The structural aspect is perhaps even more inevitable and substantial. The symbols that have a different semantic value (that represent essential traits or variable characteristics of modalities or relational aspects) cannot be interchangeable in the structure of the symbolic system: their relationships can only follow a certain number of schemes. Symbolic representation of any aspect of reality requires that there be at least one symbol that represents the essence of an object or relationship (a symbol with the function of a noun); qualifying symbols may only be associated to other symbols, which may also have a qualifying function but in the end must more or less directly refer to a noun; symbols that represent a relationship (verbal forms) constitute the most relevant semantic aspect of symbolic representation, and since most relationships are characterized by asymmetry they must be located with respect to other symbols (or modified in their form, conjugation and suffixes) according to precise rules that let you identify in an unambiguous way the direction of the relationship (who is the subject and who the object of an action, for example). Symbols that represent modalities (like the adverbial forms in any language) are typically dispensable − they only add information − but specific rules will determine which other symbols they have to be put in relation to.

All this brings us to an observation that is initially intriguing: whatever be the symbolic instrument that will be employed to represent and interpret reality, and possibly to analyze it abstractly or reproduce it, or to communicate one’s own reading, given that the symbol system is generated by a neuronal network with the characteristics of human brain, it will necessarily display some features of morphological variation and syntactical-grammatical structure that are fixed and shared with any other symbolic system that has been generated by a similar neuronal network. In other words, all human natural languages must share some structural features and traits of morphological variability. This has to be true because of the mechanism by which input information is decomposed and analyzed, translated into a neuronal activation scheme that constitutes the composition of all specific and general aspects that can be detected in the sensory experience: elements, characteristics, relationships and modalities.

As usual, this may be a curious observation, but it exhibits little originality. Nothing new in this. The analysis of human natural languages, rigorously undertaken by Noam Chomsky, has already given us this reading frame, extraordinarily powerful and capable of unifying traditionally scientific subjects with fundamental philosophical questions. A part from the critiques that any attempt at re-reading is exposed to − and many more there are, and less relevant to the substance of the argument, if the nucleus of the underlying intuition is really revolutionary − nobody can deny that Chomsky’s discoveries have framed with extreme clarity and on new bases any modern attempt at facing the topic of natural languages. His fundamental intuition, solidly supported by evidence, is that natural languages, although they show spectacular phonetic and terminological differences, are linked by strong analogies in terms of structure, by common fundamental grammatical characteristics, that indicate that a major fraction of the structure of any natural language must be based on biological, genetic, neurological properties of man.

So, we find ourselves wandering even farther. Information enters the brain, branches and disintegrates into a myriad of representations and this way triggers an abstract ballet that justifies abstract logics and even language. We must come back to ground, sooner or later, if we want to discuss what rules our behavioral response.

But it is still too soon. We have not talked about emotions, yet. We have not even named conscience.

Later. We shall think later about the way one acts. Let’s think, before acting!