The ancient Egyptians mummified the dead to preserve the physical body. The body was needed in the Afterworld. While the Egyptians seemed to have concept of the animating facility as provided by a soul, the soul for them resided in the heart, famously weighed against a feather, to determine if the heart was freighted with evil to preclude an afterlife. Still for the ancient Egyptians heart and brain were both unique organs of the body, though not as they are for us. The heart was unique in that no other organ was left inside the body. All other organs were removed to be preserved in canopic jars,  except for the brain, which was uniquely sucked out and thrown away. Possibly the brain was felt to be a substance whose only purpose was to fill the head. Aristotle also underestimated the brain in thinking it merely served to cool the blood, and that the heart was the seat of sensation. This might be considered Aristotle’s greatest error as his genius was most best expressed in his Biology which as son of a physician, he helped establish as worthwhile to study.

One of the first to recommend the brain as the seat of intellect and thought was Alcmaeon of Croton, possibly a Pythagorean, who believed in dissection and described, among other things, the optic chasm by tracking the optic nerves behind the eyeballs. Given the presence of sense organs on the head, nose, ears, eyes, tongue situated around the brain, how people talk to each other and exchange ideas by approximating our frontal lobes and occasionally looking each other in the eyes, it amazes me that the ancients did not intuit that obviously the brain is the seat of intelligence. The Hebrews, well over 3000 years ago, placed a box on the forehead symbolizing intent and one on the left arm which was to be closet to the heart, meaning action, the tefillin used in prayer, so at least they must have had the idea which reminds me of cor et manus heart and hand award which is awarded in medical schools.

Among the first to appreciate the true complexity of the brain were the first French scientific neurologists of the 19th century post enlightenment. Every effort was made to logically correlate neurological deficits with lesions in the brain at the famous Hopital de Salpetriere.   Most famous in the 19th century were the aphasias, disturbances of language function. As the frontal lobe is mostly involved with action and movement, efferent function, lesions in the inferior frontal lobe convexity were associated with disfluencies in speech output or efferent function and named for Paul Broca. Similarly the temporal lobe of the brain processes and interprets sounds and lesions in the superior posterior temporal gyrus cause difficulties with language interpretation, termed receptive, and are named for Carl Wernicke.

Knowledge about the brain was hard-won and gathered momentum slowly only over the last 175 years or so mostly by the tedious study of lesions, and connecting lesions with deficits to establish patterns of anatomical localization which is still our major window onto the function of the brain.

Up until recent years the work of the clinical neurologist was to localize abnormalities in the nervous systems and to connect abnormalities with disease. This was done in cases of infection, stroke (infarct), degenerative, toxic, genetic processes with the aid of more sophisticated tools, especially dissection in living patients or in post-mortem material, and the microscope. The chairmen of neurological departments were pathologists, always striving to connect gross (seen with the naked eye) and microscopical lesions with diseases. In some cases the work of localizing disease was very helpful as with benign growths, abscesses or pockets of blood, which could impede the function of neurological tissues and might simply be removed. In most cases localization was an intellectual endeavor that capitalized on knowledge of neuroanatomy, but added little to the lives of persons with neurological disease. One spectacular effort in localizing function was electrical stimulation on live patients done by Wilder Penfield as discussed in my post on The Origin of Consciousness.

The central nervous system is composed of clusters of neurons performing certain sensory or motor tasks. These neurons extensively connect with each other over white matter tracts in order to bring about crosstalk and further processing of information in what is increasingly termed a connectome. Neurons abut each other too and communicate via synapses.

The nervous system may be seen as a whole which is the physical basis of a single animal’s experience, while for other scientific observations, the nervous system is best analyzed as a series of localized parts. Thus the brain scientist is a little like the nuclear physicist with his particle vs wave theories. In the case of the brain particle v wave is part v whole in the sense that some problems are best solved by considering localization of function and others by looking at the entire structure.  Brain scientists have debated endlessly about whether the brain would most profitably be studied as a whole, perhaps  a hologram, in which functions are multiply and widely represented. For instance you can’t expunge a specific recollection by just removing a volume of brain. Memories and records of life lived are widely represented all over the brain. And you can’t just wipe out a lady’s PhD by removing say a piece of her temporal lobe. The brain has evolved as a helper and repository of  personality of one single individual. For further discussion please see my description of the one head one person hypothesis, in which first century notions of possession that ended up contaminating early medicine and religion (and even believers today)  with cures via exorcisms are discredited. The brain is an organ like any other that exists FBO its owner. The understanding of parts of the brain has also been extremely useful. The brain needs to be analyzed but also seen whole.

In recent decades, rapid advances in neurosciences have caused neurological localization and pathology and lesional  neurology to lose its luster as great strides are made in radiology, chemistry, pharmacology and genetics.

I vividly recall when I first learned that the cortex of the brain was organized into functional units called columns.  This was before I had a grasp of the basic anatomy of the brain as I had not yet gone to medical school. Brodmann and others at the turn of the 19th to twentieth centuries with the aid of special stains and the microscope, noted that the human cerebral cortex was broken into up to 6 layers, most often less, of different types of cells. The largest cells with the longest axons and communicating over the longest distances, were pyramidal cells in layers 3 and 5. The most complex areas with the most cellular layers were felt to have been the last evolved and most developed, the so-called neocortex or new cortex. Cells in layers or columns related to each other in a stereotypical ways that were reflected in their various forms and connections so that you had a standard pattern of up to 6 layers which was repeated in adjacent areas hence the vertical column. The columns both interrelated but mostly tended to function as a single units. Vernon Mountcastle  emphasized the presence of these neural columns in his physiology that was part of the curriculum of every medical student. Hubel and Wiesel took these ideas further finding individual neural respond to specifically localized and situated stimuli in the visual system. Individual neurons had their own receptive fields. For their work they were awarded the Nobel Prize in 1981.   Basic sensory processing was the function of columns of cells in the cortex of the brain now thought to be 0.5 mm in diameter and about 2 mm deep.  Columns of cells might change their electrical activity to visual stimuli from only one part of the visual field, adjacent columns responding to adjacent parts of the visual field.  Hence a visual sphere of the external world was mapped in pixillations onto the primary visual cortex in the occipital lobe of the brain. Other columns of cells responded to more abstract visual patterns say a line in certain orientation, so that the visual task was divided or broken up into smaller subtasks within the whole of the brain’s cortex.  The cortical column is the basic anatomic and functional unit in the cerebral cortex. Cataloging changes in cellular architecture, Brodmann and others finally recognized and identify some 52 separate “Brodmann Areas” under the microscope. Architectural that is microscopical visual relations between cells forming columns, reflected elemental physiologic function, not only for vision, but for other sensory modalities and even motor function. I remember jumping to the conclusion that the external world was directly mapped onto the cortex of the brain and thinking that this must be one of the most beautiful and useful concepts I had ever learned about.

This is a wonderful concept when you think about it. You have a group of columns subserving similar functions in adjacent patches of cortex which may be further grouped into larger and larger areas until you are able to construct a boundary between areas ruled by different architecture between cells. You can examine how neurons relate by looking at structural relations from the superficial down to 6 layers deep. As a general rule, function tends to follow structure. A good example is primary visual cortex studied by Hubel and Wiesel. On it you have the whole external visual scene mapped out in pixillated fashion, the top part of the visual field going to the bottom, the bottom to top, and the right field going to the left, left to the right purely because of the way light crosses in the eye through a lens to the opposite part of the retinal screen. The primary visual area, V1 is at the back of the brain bordering the inter hemispheric fissure. Further forward are other visual areas that handle images somewhat more abstractly, correlating form as reflected in the architecture of cellular relations. which will successively be called V2 and V3. Here the visual space is mapped directly onto patches of cortex. Other sensory modalities function in exactly the same way. We have the primary sensory cortex on which body parts are mapped onto the brain and motor areas too mapped in the same way. Much of the cerebral cortex is occupied by these one to one spacial mappings, the primary simple projection areas being flanked by secondary areas that take on more and more abstract processing. Finally there are separate regions of cortex that are multi-modal, correlating whole experiences which are our memories visual, auditory, olfactory of whole events extensively processed and stored by the brain.

Realizing the functional utility of cerebral columns which go from superficial to deep over the cortex there is an active strategy to study these and hopefully one day to reproduce the localized columnar function of the brain in a computer model. Computer scientists dream of one day reproducing the function of the brain. This is AI or artificial intelligence. One essential strategy for developing AI is reverse engineering. One can capitalize on deep knowledge of the brain to design equations and circuitry to simulate brain function. The cortex of the brain contains inner projections of the visual, auditory and other sensory worlds and a large number of inner projections of both exact models of our brain’s owner’s body and the external world. Not only patches of the visual world but body parts, even individual auditory tones and smells have certain cells and certain columns of cells that respond only to just them. Everything external is multiply mapped onto an internal cerebral lattice. But the molecular structure of the brain is in columns of cells. The brain has within it innumerable inner facsimiles projected from the owner’s body and outer world, inner little persons or homunculi and models of the world. These may be manipulated at will in one’s imagination without having to suffer the dire consequences of the the real world.

An idea for creating a computer brain model is studying and reproducing the columnar structure of the cortex. The most famous proponent who has been studiously been working on this since the 1980s is Henry Markram chief of the the Blue Brain Project, after IBM’s Deep Blue supercomputer project most famous for challenging humans in chess but the idea is brute force, harnessing a large enough computer to simulate the wondrous complexity of a human brain. His idea is to study and simulate the rat column and then the human column. If a rat column may have 10,000 neurons and 10^8 synapses a human column might have 60,000 cells and a human brain is about the size of 1000 rat brains. By 2005 it was felt a single cell computer model was complete by 2008 the first rat 10,000 cell rat column was simulated, by 2011 cellular mesocircuit of 100 neocortical columns with 1M cells. By 2014 a cellular rat brain facsimile of 100 mesocortical circuits totaling 100M cells was made. On this quantitative basis progress was palpable.

Once you have these computer models, you are free to ask as many questions as you like and you don’t have to worry about living subjects. The dream is that you might be able to simulate human or even super human intelligence. From Markram’s famous 2009 Oxford TED talk widely available for viewing, it is clear that he fully understands about the uniqueness which is the basis of all biology not only of individual animals and humans but also about diversity at the level of the single cell, that each neuron is unique. Non-living machines reproduce identical parts, biological entities make non-identical replicates, that is one fundamental difference between the world of the living and non-living. On the other hand, it should be possible to develop a standard set of rules, mathematical equations, based on relationships between cells, to understand for example the peculiar summations of functions through synaptic connections that influences individual neurons to alter voltages and the decision processes, for example the decision to fire or not to fire given the presence of up to thousands of synapses on one cell. Even then you are likely to vastly underestimate the effect of connections between regions of brain through distant white matter tracts what is called the connectome. There is a vast modern literature on mapping and following white matter tracts in the center of the brain in living subjects, what is called tractography with its explosion in radiologic studies and techniques. It seems the influence of these more far reaching connections may be underestimated by the above methods.

Brodmann discovered how intercellular architecture differed all the way at the turn of the 19th to 20th century and was able to surmise that architectural difference makes physiological function. The brain, like other organs, is composed of units of various types, neurons organized in groups, beautifully seen in cortical columns but also the glomeruli of the cerebellum centering on Purkinje cells. These are units of structure and function. Being biological units they are non-identical. That means we need to be careful in constructing physical models or at least need to be cognizant that individual components are not identical. That may complicate constructing composite models of function. Still there is lot to learn about heuristics, rules of the road, especially in forming equations and computer models.

To the best of our knowledge the brain evolved in empirical space to increase fitness the animals who own it.  I say empirical as it evolved and developed too, in an intense environment of trial and error, success and failure. For instance the occipital visual  brain of a person who becomes blind in early childhood is very different from the normal sighted individual. Other modalities particularly auditory function, may be reassigned to make a musical genius.

A computer brain simulacrum must have access to empirical space in the same way. In other words its circuits will need to have the ability to learn in the real world as a real brain does. The computer chess master may have memorized all the games, but will eventually have to improve by playing some real chess. Designing a computer brain on the basis of understanding of the biological form would seem wise. I can’t say whether the idea of building it from the ground up on the basis of understanding of biological units of function such as columns might be the most efficient approach to this problem.

One thing that fascinates me about the brain is how strongly it is tied physically to its single living owner. I can’t help but think that a computer, which lacks a dedicated possessor, will ever have a unifying influence of the self that it takes to form a conscious entity, life in other words.  I may be wrong.