My previous essay on Carbon and Silicon, I hoped, turned round the oft-asked question about whether computers are like brains. I worried more about human cyborgs and people forced to act like computers and robots. Are human (and animal) brains really just complex computers? Ifso then simply as a matter of time, computers and robots will overtake us.

I see this already happening in my daily work in the hospital. And as you can see from the following paragraphs I am really of divided mind on this topic.  Doctors are told not to think on their own and we are told not to teach our young medical students and residents to think. If you think you will get into trouble. Just about all actions have been taken over by protocol, sequential models of optimal performance.  The advantage is that a certain diagnosis is less likely to be missed. Surgical procedures are less likely to have important steps left out. Doctors, it is argued, always think their next case is like their last, because the last case is what they remember best. Clinicians who think the most, may make the most mistakes. Better to follow a protocol, test everything and miss nothing. Follow the procedure  to the letter and never leave a surgical object in the patient, or operate on the wrong side of the patient. But those of us who have practiced for a long time, wonder about surrendering our frontal lobes which we have honed over these long years at the gates of the hospital. I like to say, as in Dante’s Inferno, “Surrender all hope”, we doctors are told “surrender your frontal lobes” all who enter here. The practice of medicine has thus become wooden, robotic, dehumanized. I strongly suspect this protocol driven practice extends much farther than in of medicine to most other fields.

The study of the brain has always been based in anatomy, especially clinicopathological correlation which localized specic functions to certain regions of the brain. The nineteenth century saw the flowering of anatomical localization in the brain which rested almost entirely on brain dissections done to correlate brain lesions with deficits. Diseases of the brain were understood to pursue a certain clinical course and involve specific regions of brain. For example Alois Alzheimer found that certain microscopical changes occurred first in the hippocampus, a region implicated in recent memory function, and later more severe microscopical changes accumulated as the disease progressed. Other neurological diseases such as multiple sclerosis and poliomyelitis were similarly described according to microscopical and anatomical changes that could be clinically correlated (pathology). Thus the model that developed at the time of brain function was primarily lesional, that is lesions or diseases induced neurological deficits.  A specific lesion is connected with a deficit allowing one to draw conclusions about the function of a particular region of the brain.  For example Broca connected a lesion in the Left inferior frontal lobe with dysfluency of speech that often accompanied right sided weakness. More precise descriptions of language dysfunctions, termed aphasias, were more precisely localized strictly on the basis of their lesions. Popular descriptions of cases such as that of Phineas Gage who suffered a penetrating wound to his frontal lobe would be one of a huge library of examples.

As brain function was correlated with thought and mental function it gradually became possible to lay to rest superstitious notions regarding epilepsy as possession.  The best example is First Century New Testament accounts of epileptics via exorcism.  For the first time epileptic events were connected with lesions and diseases of the brain. In the twentieth century devices came online whose purpose was to localize brain functions and it became possible as never before to follow non-invasively the electrical activity of the brain which was found like heart, muscle, and nerve to be constantly electrically active.  Hans Berger invented the EEG, modeled on the EKG, that enabled monitoring of the living human brains, especially valuable to follow levels of consciousness and define epileptic seizures, seen for the first time to be electrical events. Drugs were developed  and found to work mostly by controlling abnormal electrical discharges. In the 1940s Wilder Penfield tried stimulating the cortex of the brain during awake neurosurgery and thus mapped the motor and sensory cortex of the brain. It was as if a little deformed man or homunculus was doubly draped over the convexity of the brain. Penfield obtained this data using electrical stimulation of the brain.

An explosion of knowledge in recent decades owes much to the computer. Rapid computation has revolutionized brain imaging. X-ray exams over my early career as a neurologist allowed imaging mostly of the skull and spine whose Calcium laden bone attenuated the x-ray beam. We did use contrast agents which could be injected in the spine, yielding columns of dye, or in the blood vessels creating blood vessel pictures called arteriograms. One particularly cruel and unusual test, often caused dizziness, and vomiting. The  pneumoencephalogram involved injecting ordinary room air into the spinal column through a spinal needle and then turning and tilting the patient in hopes of following the dark air bubbles around through the surfaces of the brain and into the inner cavities called ventricles.  That primitive test was in use until  just about when I started medical school.  In all of these earlier x-ray techniques, clinicians were unable to see the brain directly, only the deformation of dye columns, air or blood vessels that might be pushed aside by tumors, injuries or pockets of blood or injuries. Actual brain anatomy was a matter of inference as the brain and spinal cord were not viewed directly.  Therefore  It took a great deal of knowledge to come up with an accurate diagnosis.  Diagnosis was far less explicit or reliable, no matter what the skill of the observer. I started practice at the tail end of this age which I refer to jokingly now (but tragically for patients at the time)  as when men were men.

It is only in recent decades, the CT scan was in use after about 1974, the MRI 1984, that clinicians were able to see the brain and spinal cord directly. That was revolutionary. And it came as a direct result of computation. With CT, computerized tomography, x-ray beams were partially attenuated, obtained and stored in multiple short moving shots through a given region then rapidly integrated to form a coherent image by computation. MRI used similar techniques but no actual radiation. In our atomic age, you could take protons and make them wobble by subjecting them to a magnetic field, watch them resonating and then integrate these signals through rapid computation to make spectacular images. You can see nervous structures, to me a thing of beauty, directly, as never before. All manner of pathologies could now be viewed without the aid of a scalpel or a microscope and it was possible to make localizations as never before. There is nothing like a direct view.

Computers have allowed us to understand anatomy in a way dissection never could and disease in the four dimensions of time and space. Physicians still maintain their anatomical notions about neuropathology. The physiology or working of the brain in health and disease can be seen in living creatures as never before and without invading the body. Specific functions and activities of various brain regions can be watched as on functional MRI and PET scanning which compare changes in metabolic activity over time of parts of the brain in persons for example in the midst of a seizure or imagining scenes or working out a problem.

Computers have given us more than that. Computers have begun to simulate higher brain functions, solving arithmetic and mathematical problems and at great speed, storing huge quantities of data in magnetic and other media, creating images, doing linguistic and musical tasks, functioning as precision robots in automotive and other factories. The computer is a simulacrum or mechanical doppelgänger of the mind. Gradually,  those of us who have been brought up with a biological model of the brain, steeped all of our lives in anatomy and physiology, have had to accept a computational model of brain function. Yes, there is a gradual shift from an anatomical model to a computational model of functioning of the brain which I have seen in my lifetime.

One useful example is a modular view of brain functioning. Everyone knows there is gray matter in the brain composed of the cell bodies of neurons which looks gray, and white matter which is from myelin, that connects gray matter regions termed in the central nervous system, nuclei.  The white matter consists of cable like axon connections termed tracts. Small groups of cell bodies are now conceived to perform specific relatively homogenous functions and thus are best conceived as sort of plug and play modules. For example, Broca’s area, mentioned above, puts words together to make phrases then sentences. Silos of of words are stored elsewhere and accessed via connecting axon fiber tracts for use in verbal expression. Thus neurological tasks or deficits are now conceived as some  combination of modular functions and connections between modules.

For example after a severe brain injury many patients languish in a state between coma and awareness along a spectrum of disorders that ranges through coma to persistent vegetative state, to minimally conscious state,through to partial awareness of one’s environment to fully cognizant state. In any given patient at any given time, certain regions of cortex may be active or inactive, each of these cortical regions is now looked upon as a module. The fully aware patient is able to recruit a given a full complement of cortical modules at will while the less responsive or minimally conscious individual may recruit less, the persistent vegetative state even less. Content of consciousness may be compared to a certain number of lights being on in a skyscraper, more than a certain number of lights (modules) equals higher level of consciousness. In addition we have switches controlling the activity of cortex in general, residing in the brainstem, which determine the general level of arousal of the brain heretofore called the reticular activating system. These affect consciousness along an axis of arousal ranging from unarousable unresponsiveness, to  obtundation , to stupor, lethargy to behavioral  wakefulness whether or not the requisite number of cortical modules is activated. So we can break the notion of consciousness into two axes: 1.  awakeness or arousal dictated by the brainstem arousal mechanisms and 2. content of consciousness which is determined by activity of cortical modules higher up in the brain.

Brain modules  may in addition exert countervailing influences, mostly via inhibition of other brain areas but sometimes through augmentation as well. Attention is the assignment of activity to certain brain modules. Greater effort entails enlisting  or recruiting wider and wider brain regions. These changes can be explicitly viewed on modern tests such as functional MRI and PET scanning. The modular view of brain function which is all about modules and their connections, yields a more precise appreciation of the brain

The next step, already well underway, will involve the interaction of living, Carbon based devices, with mechanical primarily Silicon devices in direct proximity to nervous tissue, what is termed the marriage of Carbon and Silicon. Limited Silicon based devices may be implanted within the brain, the most basic of these are merely oscillators which deliver periodic electrical currents, but more advanced devices are also being developed to aid in acquisition of sensory such as auditory or visual data and implanted directly in the skull in essence plug and play devices.

In summary, computational devices have transformed our view of our own inner selves, from a primarily biological or anatomical model to a computational model. Thus the terminology has changed. New concepts carry deep insights about mental function. Gray matter, white matter, have been replaced respectively by modules, and the connectome. Classic neuroanatomy will count less than than connections made between functional modules.  The brain may well be redesigned by abundant devices likely to be invented over the next few decades. Human mental capacities especially sensory functions and memory, are likely to be enhanced immeasurably and may be altered in other ways.

One troubling question is whether humans may thus come to resemble computers, or robots in their increasing mechanization, becoming wooden, robotic response machines. Will humanness whatever that is, win out or even be enhanced or otherwise might our behavior come to resemble machines in their reliance upon algorithms protocols and formulas. What is in the future, if any, for exquisitely human characteristics such as emotion, delight, spontaneity, plasticity,  imagination, invention?