Main Publications Our Practice Links Feedback Events
Main Publications Our Practice Links Feedback Events

 

From Yanofsky’s BEYOND BIOLOGY:  INSIDE THE NEURON (Chapter I) Continued

Part  II
 
Neuron as Executive

As we see the nervous system as a whole in its executive function, a central command and control station, then we conceive of  the individual neuron as a single miniature unit in this decision making process.  The neuron or single nerve cell is a mini-executive, a single soldier in an army of neurons.  Nerve cells have different rank, depending on how exactly they connect with other neurons, how many other cells they connect with, and their specific location.  But the analogy goes only so far, since there is no one cell generalissimo in the brain, rather a complex feltwork of cells.  Each cell is subject to a complex of inputs. In most cases the individual neuron makes a binary decision, that is, at a given moment it is either firing or not, similar to a series of 0's and 1's in a computer. The computer analogy is even more obvious but even more imperfect, as we shall see.

If the nervous system is a complex of individual cells each of which at any moment is in one of two states, excited or not, the brain is nothing more than a large scale information processor of complex binary digits (bits). Understanding nervous function, hence thought and feeling, comes from data about the either on or off state of each neuron, and a knowledge of how these neurons are wired together (their anatomy). The oft used highly seductive brain computer analogy has, until recently, been very compelling and has a number of attractions, not the least of which is the resemblance of a logic machine to its maker which is the brain.  In this computer logic machine, the brain sees a reflection of itself. A computer may think and if it can its thought processes must be much like the men who creat­ed it and may even be used to model for human thought processes.

 

In the computer, you have the human brain creating a functional machine in its own image. Therefore brain scientists may learn a great deal from looking at computers and logic circuits.  Many aspects of neural function can be understood from the vantage point of computer design. To mention just one of many examples, we tend to look at brain function today as a composite of function involving modules, circuit boards essentially. Consciousness, as we have seen, involves various groups of cells, each group responsible for a specific function. The ARAS awakens the brain, the cortex gives content to consciousness, the emotional or limbic centers color consciousness, the frontal lobes as motivation and will, and so forth.  Take out a module, say the limbic system, and consciousness will be altered, perhaps the subject will be awake and aware, but will lack emotion, will be a Mr. Spock or Data. The more complex the brain and the specific function you are examining, the more useful the concept of modularity.  A group of neurons is a circuit board or even a chip.

 

Computer engineers have learned a great deal from looking at the brain as well.  They are constantly trying to emulate brain function and are hungry for knowledge about brain circuitry.  Parallelism in brain function was not appreciated until fairly recently.  Your ordinary computer has a single microprocessor.  All information must flow through this single unit which functions at great speed, but still, every piece of data must flow through the microprocessor in sequence, one bit at a time. Logic is thus 100% sequential.  The brain however, is capable of handling a lots pieces of information in parallel, through many pathways simultaneously.  Each individual datum in the brain is processed by computational standards, fairly slowly.  In order to walk we must simultaneously and in parallel process three dimensional visual data, input from vestibular balance centers, proprioceptive input from our peripheral nerves to mention just a few information sources, then compute a whole program of muscle activations and then vary all of this over time, all this just to walk.  No wonder computers have not up until this point been able to turn out anything resembling a full human gait pattern. Sensory input in the brain and motor output, are “massively parallel”.  Innumerable microprocessors, that is, neurons and groups of neurons, are working in the brain, in parallel.  In recent years, computer scientists, realizing the advantages of this major divergence in design, have built it into computers, which now boast vast arrays of microprocessors working in parallel.   

Computers are growing more and more powerful, partly owing to designs that emulate brain function, but far outstripping human capacities in many areas especially data storage and retrieval and in calculating ability. It makes you wonder whether computers will one day be able to “think” and whether output of advanced computers, say computer speech, literature, or musical composition will always be as inferior as it is today from the output of talented humans.  Do we expect computers to become high-speed initiators of perception, thought, feeling, and action, be conscious in other words?   Or computers may even best people in some intellectual endeavors.  Is it possible to build into a computer self-awareness, anxiety about death, an idea of personal boundaries and space, emotion, all the things that define awareness of biological Carbon based beings?  Time will tell. Superficially computers, mechanical objects based on Silicon, resemble Carbon based biological machines. (Of course, we have created a cognitive tool in our own image!!) 

As neurons exist in two states only, there is an electrical action potential or there is no action potential, so computer function can be reduced to a series of 0's and 1's that is translated on higher and higher abstract levels until with all the things computers do, word processing, mathematical reasoning, switching and other tasks this series of 0's and 1's is invisible. Brains and computers are both machines, so the argument goes, and the basic structure of each is not relevant to the similarities in their behavioral output.  Carbon and Silicon are enough alike in any event and each machine, brain and computer utilizes arrays of binary elements that at any given moment in time are in one of just two states.   The precise configuration of binary elements, invisible when you are looking at a specific task or output, determines content.  This brain-computer analogy is at once very seductive and false, seductive because it works and is actually a good description to a limited extent.

First, the binary analogy is not strictly true for neurons.   Though at any given moment a neuron is either firing or not, and hence may mathematically be assigned a '1' or a '0' to describe its state, this is nearly always an incomplete description of the neuron's state at a given time.  The '1' for the firing neuron is straightforward enough but after firing almost all neurons will be unable to fire again for a certain length of time, which is the refractory period for that cell.  The nerve cell has an absolute refractory period over which time it cannot be made to fire under any circumstances, and a relative refractory period when it is merely more difficult to excite the neuron.  A given person, is only able to achieve an orgasm at a certain frequency.  For a time you may try to excite him and get no response at all.  Over a longer period of time, it is difficult to excite him though that can be done. The neural action potential is similar to an orgasm. There are the absolute and relative refractory periods. The refractory period for neurons is different for each cell. 

Further, at a given moment, a neuron is at a certain state of excitation.  A mathematical description that assigns a single value to a neuron dependent on whether it is firing is not completely describing that cell. Superficially neurons resemble Silicon based machines whose states may be represented as a series of "all or none", firing or not firing,  '0's' and ‘1’s’, but mathematical models may not take into account that at any given moment a neuron is more or less easy to get into its firing state, it is more or less excitable.  We shall see that excitability can partly or wholly described by the current state of depolarization of the neuron.  The neuron responds by adding up in some way all of its inputs, typically synapses, excitatory and inhibitory, which can run into the thousands for an individual neuron.  If it has only just fired, it can't fire again.  In addition, there are global factors that modify neuronal excitability including among many things, fatigue, hormone and drug effects, availability of energy, and Sleep/wake State, appetite, and satiety.   For example, an animal that hasn't eaten in a long time is hungry and we may say accurately though in a simple-minded fashion, that neurons in a putative hunger center are poised for action.        

Not all neurons are locked into a binary paradigm anyway. Some make a graded electrical response.  Some of the light receptor cells within the retina of the eye function in this way, but also many other nerve cells particularly sensory cells.  Nerve cells store data in a number of ways.  Neurons that we know about are affected by elec­tric charge in individual cells, but another possibility is to change in electrochemical relationships with neighboring cells. There is good evidence that changes within the nervous system are induced internally not only via the acquisition of new informa­tion (learning or change in software) expressed in changes in electric and chemical signals, but also in changes in structure. Nerve cells and synaptic connections between them form and break down over time. There are also permanent changes within nerve cells that occur with environmental change and in learning. For example, it is thought that learning may involve synthesis of intracellular chemicals such as RNA.  Unlike the array of binary electrical elements within a computer, neurons are influenced by supporting cells especially changes in other numerous cells called glia (for "glue", the cells that help bind neurons together and perform myriad other functions).

 

EVEN ON A CHESSBOARD:

Nothing illustrates the differences between the human brain and computer as well as the game of chess. It should be emphasized that for most fields of endeavor the computer, even for modern machines, is so much more primitive than the brain there is no means for comparison.  But chess is an artificial situation in which moves are confined to an eight by eight square board and pieces having restricted and well-defined geometric moves in only 2  dimensions, a task  perfect, or so it would seem, to pit a calculating machine against a human brain.  In real life we work in three dimensions and consider many divergent inputs at once,  but even in the limited field of chess, profound differences can be seen between the workings of a conscious machine which is the human brain and a computer.

You don't have to know much about the game of chess to appreciate a difference but it helps. Chess is a game of strategy that involves analyzing one's position and a specific sequence of moves that needs to be accomplished to maximize future positions. It is meant to simulate a battlefield except that it is, eminently sequential.  A real general needs to manage a number of changes occurring at the same time,  Not so on a chessboard where each move occurs in sequence,  hence chess should theoretically be a  again a perfect task for a computer since moves are, by nature, sequential.  The chessboard has become the standard field to pit brain power against computer power.  It's great too because it shows fundamental differences in action between brain and machine. 

A human plays the game by developing a strategy and pursuing it. The human chessplayer is a schemer capable of acting on a number of hunches at once.   He looks at his position on the board, sees an opportunity and makes a plan to accomplish a perceived goal. In short, he strategizes. He may very well know his opponent and his weaknesses but the most important element is that a person keeps a specific goal in mind and makes plans on how to attain it. He may covet control of a specific space or region, say at the center or the board or may need to capture a certain piece.  However a computer is a machine and so has no goals or plans.  For a computer chess is a particular situation translated into numbers and values.  To plan its next move it will have to evaluate its position and have some means of comparing the relative advantages and disadvantages or outcomes of all of the possibilities for subsequent moves.  Each of the possibilities is assigned a relative value which and these numbers are compared, determining the next move, so that a souped up calculation is made to simulate human behavior.  The computer is a calculating machine that compares relative values nothing more, but has to look like it is "deciding" on a move, which it isn't, what it is really doing is picking an alternative based on a numerical comparison of an outcome.  The computer looks at the possibilities for a next move, comparing, depending on its size and power, what may happen one, two, three or more moves into the future.   The bigger the computer the more calculations it can do in one second,  the more moves into the future it can compare and it can look at an enormous number of alternatives which no human can do..  The bigger the computer, the faster its microprocessors, the more microprocessors there are,  the more calculations it can do, the more moves it may compare, and the farther it can look into the future, the better human chess player it can beat. You can build a huge computation machine that will have all the advantage against a human player.   Put in a load of microcircuits and perfect the software with appropriate input from the best players.  The idea of using a big machine with rapid as possible calculation is called brute force and as things now stand it is the major method by which calculating computing machines compete against humans.   And the machine won't make any stupid human errors either. 

The interesting thing is that while computers can easily defeat human chess players of average ability, at this point they are not better than the best human players, the grand masters.   In February of 1996 there was a well-publicized tournament between "Deep Blue" an IBM machine and the grand-master Garry Kasparov.  Kasparov lost the first game, but he later won the tournament not through his incredible calculating ability, the machine was far faster than he was, but by strategizing, perceiving the weaknesses of his computer opponent, learning about the methods used and where they fell through.  It is significant that he defeated the machine only later on and not in the first game that he had to learn about the weaknesses of play, that is that there was no conscious strategy at all for the machine only numerical comparisons, hence no design, no goal in mind and method to attain it.   This concept is akin to motor planning, scheming and goal directed behavior which behaviors in neurology are attributed to the frontal lobes of the brain, one of the major factors that makes us human. You can't talk about a computer having tactics, or plans, not at least at this stage of the game, even on the limited eight by eight field of chess let alone in real life.

 

 

 

 

 

The final denouement is that in May1997 Kasparov was narrowly defeated by Deep Blue after each side won a single game (there were also three draws). This is a very close record. Some say that Kasparov at one point lost his nerve and conceded in one game before he was sure he’d lose. The computer had no nerve at all. In fact the major advantage the machine has, is the lack of psychological weakness and doubt that we are all subject to.  The computer doesn’t make silly mistakes and isn't influenced by sickness, or mood.   It will capitalize relentlessly on the mistakes of its human opponent.

Deep blue ended up winning its 6 game rematch with Kasparov, not by much, with a score of 2.5 for Kasparov, vs. 3.5 for the machine.  There were three draws with Kasparov winning the first game of the set, the computer the second and sixth.  The design of the IBM computer, the RS/6000 was ‘massively parallel’ meaning there was an array of microprocessors capable of analyzing bits of information simultaneously, not only sequentially as would have to occur if there were a single processor.  Another thing the computer had in its favor was brute force, deploying tremendous computational power.  The RS/6000 IBM machine could examine 200 million movers per second, which is a little faster than Kasparov’s brain.  Moreover the machine was able to make up for its lack of strategy because built into its design were algorithms specifically made to defeat Kasparov himself, to capitalize on his weakness. Could the same machine defeat any other chess grand master?

The human vs. the computer in chess brings up the same arguments as were raised in man’s competition with bacteria and insects.   Does the human brain which can learn and perform acts by volition, design and the deliberate strategy, have the advantage over a machine capable of examining millions of possibilities but without an aim or a goal?  Bacteria and insects can mutate many orders of magnitude faster than man, exploring millions of possibilities, though quite stupidly, along the way.  Man doesn’t muddle as much. He designs, but at the end, one wonders if the action by design is any better than the brute force exploration of all the myriad possibilities.  Which method is ultimately more adaptive, strategy, design, intelligence or brute force?  Will man win against arch bio-rivals bacteria and insects? Or you can even generalize and ask whether our world or the cosmos developed merely from the exploration of myriad possibilities or is everything the culmination of some design.  The jury is still out. 

 


 

Figure 1: Which will win out, the brute force examination of myriad possibilities, or the grand design?
 
 

 

 


In a sense the computer is a high-speed exploiter of possibilities though its process is aimless.  A computer has no goal of its own, unless a human gives it a goal, no strategy. The computer's inability to function as competently as the human brain in less structured areas that is as its own expert has been a major disappointment.  Artificial intelligence has been discussed for a very long time, and for a long time seemed within reach.   There were high hopes within the field just a few years ago. Now the term is anathema as the fondest hopes and dreams have failed to become to fruition.  Origi­nally computers promised to replace humans in vast areas of intellectual endeavor. Developers of systems of artificial intelligence would, it was thought, create human reason machines that would replace highly skilled people in various fields. In medicine computers would replace experienced clinicians, and make unerring diagnoses. At the very least they would have access to huge banks of data and would not be subject to human error. Machines could take the drudgery out of diagnosis and make it more precise. Engineers, architects, and attorneys would become similarly obsolete.

As early as 1963 Joseph Weizenbaum at M.I.T., introduced a program called Eliza,  designed to simulate questions of a psychiatrist. Eliza was not about to replace your analyst but was tongue in cheek, more of an intellectual sleight of hand. The program merely re-questioned the human subject utilizing key verbs and nouns culled from his previous response. "How are you feeling today."; "I'm feeling down."; "Can you tell me more about why you're feeling down today." This program had been able to fool human subjects who did not recognize that there was no cognition driving the programs responses, only the return of phrases in the form of questions. The quest for computerized simulation of human thought processes has turned out to be a giant disappointment, as early workers in the field made a series of promises that could not be kept.  Major strategies involved what are known as rule based systems. The idea was to compile a set of rules used in reasoning within certain fields.  Next you mix in a large data base to utilize these rules (computers are particularly excel­lent at storing and retrieving huge quanta of information) and create an algorithm or set procedure for following these rules to recreate the mind of an expert. In medicine a matching system may attach various numerical weights to symptoms and physical findings using programmed rules to determine a diagnosis. One can even alter the numerical weights assigned to these rules in order first to reproduce the competence of human experts and later to outperform them adjusting these values according to the success the program experiences when confronted with certain specific situations. Modifying the numerical value assignments of various characteristics is but one way a program can be made to “learn” or alter itself in order to be made to perform as well as or better than an expert.

However, while they have been useful as educational tools such systems have mostly failed to replace expert opinion. One famous example called "Mycin" attempted to diagnose bacterial infections on the basis of data presented. Unbeknownst to the creators of such systems, human experts don’t usually follow a given set of rules. We may teach our medical students to follow rules, and there is a certain basis of factual knowledge that is necessary if we are to perform an expert func­tion. But there is also a stage if a human is to function on anything but the most rudimentary level, where rules become expendable and are no longer followed by practition­ers. Experts function differently than beginners in that they depart from the program when called to do so. A student masters a subject by learning rules only to discov­er later that his older colleagues function at an even higher level by not following these rules precisely. Indeed, the more advanced practitioner may have cast away many rules that he depended on to learn the ropes at an earlier time in his career.

I often find medical students and residents baffled by a therapeutic course set by an experienced staff member.  The experienced practitioner may not follow the rules in any precise manner but does better by the patient. In teaching examination of the patient we always have students go through a certain sequence which they initially follow closely so as not to "miss" any specific part.  However all experienced clinicians find themselves looking at many aspects of the patients simultaneously, and if we're worth our salt, honing in on the particular problem(s) confronting us. In this way one misses very little, but an awful lot of irrelevant information is relegated to the scrap heap and may not be worth mentioning. You need to know what is and what is not important.  In outlining for the novice precise rules utilized in recognizing disease, certain findings invariably mean more while other critical elements while usually noted, are not fully appreciated. In many instances an experienced person can make a diagnosis almost instantly without resorting to rules at all.  A certain pattern of speech may make to diagnosis of amyotrophic lateral sclerosis almost unmistakable. Similarly an experienced mechanic knows before he opens an engine that a certain tapping sound comes from a loose valve cover.  This is the facility of recognition which computers are not as competent at as humans.  Computers follow rules exquisitely well, much better than medical students, but recognize faces and sound patterns not as well.   Recognition is just one of the cognitive facilities that humans have and can call upon at any time, that computers first have to be designed to have.  Then computers or any similar cognitive device would of course be expected to call upon that facility on an as needed basis as humans do, a tall order.

Prosopagnosia is a curious brain disorder that says a lot about how the brain works. In prosopag­nosia there is an inability to recognize faces, those of ac­quaintances, relatives, even oneself, a simple function we take for granted. Recognition is accomplished holistically, not through analysis of individual elements in a picture. You don't recognize a friend by noting that his eyes are spaced a certain number of inches apart, that he has brown hair or a certain shaped mouth, you simply see his face and know it.  Prosopagnosia can affect the ability to distinguish objects within a general class such as a farmer recognizing each individual cow, or pick­ing out one's car in a parking lot. Such objects are not distin­guished on the 'conscious' level by analyzing individual charac­teristics. However, a victim of prosopagnosia has to depend on a conscious search among distinguishing characteristics, such as the letters on the license plate of his car or the mailbox address of this house, or even nonvisual cues. This disorder celebrated in the book, The Man Who Mistook His Wife For A Hat and in a certain opera by the same name, illustrates certain princi­ples of brain function. Firstly recognition is immediate and is not performed analytically, that is as a sequential feature-by- feature task.  An entire object is not broken into its compo­nents. Secondly, a simple brain lesion, in this case a disconnec­tion between the visual or occipital area of the brain and the recognition areas in the temporal lobes on both sides of the brain, frayed wiring, if you will, may interrupt this recognition process. For over a century there has been a debate among brain scientists between those who sought to localize particular cerebral functions in precise brain regions (the phrenologist's approach -Phrenologists attached great significance to the study of bumps on the head) and those who took a more holistic ap­proach.

Paradigmatic among disorders in which localization of function is significant are the aphasias.  These are disorders of language function localized to the dominant (usually the left) hemisphere. After long years of examining patients with localized brain lesions, strokes, head injuries, tumors, abscesses etc., neurologists discovered that disorders of language func­tion occur when an area on the left side of the brain has been affected. Moreover they found out that destructive lesions in the frontal lobe cause weakness or paralysis, while a lesion in the parietal lobe causes numbness or a problem with sensation. Because we are built in such a way that nerve fibers cross to the opposite side of the body, destruction of the left side of the brain causes paralysis or numbness on the right or opposite side of the body. Aphasias, dysfunctions of language, follow this same general scheme. Destruction of the left posterior frontal lobe will cause a problem with movement on the right side of the body and also trouble with the motor aspect of making speech and with writing i.e. an expressive aphasia. A lesion of the parietal lobe and the temporal lobe which lie behind the frontal lobe, will cause a problem both with sensation on the right body, sometimes even visual loss on the right side, and also a problem receiving understanding and interpreting speech and written language. This can now be appreciated in the living patient using various brain scans and electrophysiologic tech­niques that weren't available when these discoveries were ini­tially made. The recognition and classification of Aphasias provides one of the most reliable localizing techniques in clini­cal neurology, lending strong support to those who maintain that there are particular functions that reside in certain brain regions. Among the persons who believed in such localization was Paul Broca.   Sigmund Freud was a no less distinguished spokesman for those who maintained a more holistic approach. But the argu­ment for precise localization in the brain is not as simple as the mechanists would have us believe. After localized destruction occurs, other areas begin to assume a good part of the affected function.  Part of the recovery process seen in patients is a goal directed.  They try to recapture lost function by whatever means become available. Various supplementary areas not ordinarily used for specific functions, begin to take over function of damaged areas not only in the opposite hemisphere, as was once thought, but also on the same side as the neurological event.  That brain areas initially uninvolved in a certain activity can take over this function in a pinch has profound meaning.  Brain functions are controlled locally with certain brain areas destined to perform specific functions.  On the other hand, most brain areas are pluri-potent, while in the natural state performing a fixed function, are able to assume other functions if necessary.  This is partly how the brain repairs itself to preserve the organism.  This is also not something generally observed in machines even Silicon-based machines.    You don't find a hard disk taking over the function of the CPU or the printer doubling as a random access memory device.  But the brain is different.  In the repair process and in order to preserve function, we see such seemingly disparate areas of brain suddenly becoming active and assuming function.  In stroke patients who are trying to perform a function impaired by their stroke, say trying to moved a paralyzed right arm, you can watch as far-flung brain regions are recruited to accomplish a task.  One method is to use the PET scan, which looks at localized glucose utilization that in the intact person are not used to perform the task.  Whereas in an undamaged person you may have seen the motor strip of the left frontal lobe become active, in a stroke patient whose motor strip is non-functional you see other areas become involved. The  supplementary motor area on both sides of the brain and even the cerebellum pitch in with their effort.   These areas are all recruited in order to accomplish a certain task which is second nature to an intact person but accomplished with greater effort when impaired.  Now other areas of the brain have to become involved presumably as this damaged individual works harder to accomplish even a simple task. The brain may be the only machine that is able to jury rig itself in a pinch.  It uses what is available.  The impetus may be strong motivation to accomplish something.  Motivation is not something you can observe in man made machines.

You can also see this when a less adept person who is not damaged performs as task that is more difficult for him. For example girls for certain math tasks, boys for language tasks.   In order to perform the same function more widespread areas of the brain need to be called in any time a  task is more difficult.  A minor difference of opinion may be settled adeptly by your diplomats. If your leaders are just a little less competent you may have to call out the army.  Such considerations have now become a basic feature of rehabilitation for example after stroke and head injury.  The brain is at the same moment, localized and non-localized, holisitic. These contrary elements, localization vs. holism,  in consideration of how the brain works are as basic to neurobiology as particle vs. Wave models are to physics. They are different aspects of the same phenomenon. The brain is a whole structure composed of modules or elements.  Depending on how you ask a question you will see one or either side of its nature.

The brain is centered about performing a certain function.  Whether it be writing, or reading or moving an arm, throwing a ball,  it will work until the job is done.  In order to accomplish a task we may have to recruit brain areas not ordinarily used to perform a given function.  It's the same when you break a leg, the leg is casted and you try to walk.   You will recruit the opposite leg, your arms on crutches but you know you have to walk and get from place to place and you do it.  A machine is different.  It's designed to accomplish a task and an algorithm or sequence of moves is incorporated into its design, in order to accomplish the task.  If one part of the sequence fails the work will not get done.  There is no goal direction, only a series of instructions.   

The patient's frustration at his lack of function seems to speed his recovery.  The human patient is goal directed something we don't appreciate in a Silicon based machine.  Because the brain is plastic, it recovers function even after cell death. An inanimate machine designed for a deliberate purpose different than an organism that develops through the biological process of evolution, trial and error, in a real environment.  The designed machine suffers from the same shortcomings as a planned economy in a communist country as opposed to a capitalist or natural unplanned economy.  Economic planning does not work as well as a naturally derived economy.   The latter, is primarily goal and task rather then design directed and is bound to be more plastic.   We continuously discover how much we tend to underestimate human and animal plasticity. Plasticity is part of what defines the brain as a living tissue.  Reading and letter recognition, which comes easily for the human brain, is very difficult to design into a machine. A program may possibly recognize a precise written figure on a specified background as long as parameters of shape and size are specified with mathematical precision. Even then there are problems in recognizing these patterns in a slightly dif­ferent orientation or in different form for example, the same letter in a different person’s script or in a different slant or orientation. The computer's "perceptions" work again via analy­sis, a non-holistic approach. Images are mostly analyzed into tiny boxes (pixels or picture elements) hundreds or thousands of these making up a final form. The position of each box is mathe­matically described. The brain performs recognition functions easily because it does not work through analysis but rather per­forms its function more holistically. Computers bear little resemblance to human brains. Computers perform logical processes sequentially their responses being hard wired and determined. By contrast the human mind usually functions in a non-sequential manner. Thus a computer with one loose connection will probably cease to function. The brain loses components, nerve cells and other elements almost continuously yet still keeps on all the while improving learning and increasing func­tion, primarily because it does not depend on sequential but instead mostly parallel and overlapping processes. As a living organ it is constantly working and repairing itself. a malfunctioning or sick brain most of the time doesn't lose its oneness, personality and  basic method of coping.  A human with prosopagno­sia or aphasia is still the same human.

Comparing Brain to Computer

Computer scientists, realizing how the brain works as a parallel  machine, have tried to emulate brain function with their electronic silicon based machines. They have realized the advantages of designing machines that reflect biological methods. This means using a strategy of parallel instead of sequential processing.  In a computer everything must be processed by a centralized processing unit (termed the CPU) which is a microchip.  All operations need to wait their turn and go through this tight bottleneck  one step at a time,  which is why machines emphasize the speed of this microprocessor (for example the Intel 486 or Pentium Chip.) Experts have designed parallel distributive processors (used even for the case of the eminently sequential game of chess, by the way).  One of the tasks they have set out to mimic is recognition.  In massively parallel machines with overlapping function the loss of a few elements does not shut down the function of the whole machine.  For a serial processor the likes of which are our own home computer, the loss of any particular element, especially the central processing unit, would end everything.   Newer machines employ parallel arrays of CPU’s instead of just one the CPU being Silicon based analog of the individual neuron.  It is common knowledge that as we age, we lose tens of thousands of neurons daily, yet our performance in certain tasks, especially in our 20's, 30's and 40's actually improves as learning takes place and we see perhaps what even amounts to an increase in synaptic connections.  Sooner or later the loss of neuron's effect, overtakes the offsetting process of learning and we see the ravages of aging on cognitive function.    Only in recent years has this offsetting effect of learning on aging been fully recognized.  It is one potent method to delay aging effects. 

 

 

COMPUTER (MACHINE)

BRAIN

PROCESSING

SEQUENTIAL

PARALLEL

ATTACK

ALGORITHM

STRATEGY

PRODUCTION

BUILT

DEVELOPS

ACTION

FOLLOWS ORDERS

INITIATES ACTION

BEHAVIOR

DETERMINED

FREE WILL?

MATERIAL

INORGANIC ( SILICON)

ORGANIC (CARBON)

REPAIR

OTHER

SELF (HEALING)

DESIGN

ENGINEERED

EVOLVED,  GROWN

Table 1: How machines and brains differ.  With efforts to make computers match human attributes, these distinctions blur.

 

Obviously the brain is function­ing constantly as a correlator and user of input from numerous simultaneous sources. A football player is waiting for a pass from the quarterback, but he also has to keep alert for players of the opposing team who threaten to tackle him if a successful completion occurs. This information compared with an internal program consistent with plans for a play accentuated by practice.  Then he has to position himself to make a run for the goal post.  His vestibular system and cerebellum need to keep him upright and moving, but most importantly, he has to somehow estimate an optimal position for his body, and hands in order to successfully make a catch. Just some inputs include the timing and angle of the quarterback's release, estimated veloci­ty, his own velocity and direction, all this computed instantane­ously, and well beyond the capacity of any mechanical contrivance. In plain words the brain acts as an executive, receiving input from disparate sources and putting them together in order to accomplish certain goals.  It must process all of this multimodal internal visual auditory input and then to issue orders for a play and run to the goal post.

The brain is an associative instrument, correlating and processing in parallel input from disparate sources.   In regard to visual operations alone:  "Considering the process­ing that takes place with visual input, sequential processing would not be possible.  If it takes 500 milliseconds for a person to respond to a visual recognition test then there must be no more than 100 synaptic steps between the input and the output. Accordingly, a hypothesis that envisions a serial processing unit for visual recognition with 300 to 1000 steps cannot be right. This observation is usually followed by the inference that the brain, unlike a standard electronic computing device, is a mas­sively parallel machine The point is 100 steps in a serial processing program is far too few to do anything very fancy."[1]*

 

 

 

Marriage of Carbon and Silicon

Computer scientists are not only using biological modeling for new parallel processors.  They may incorporate carbon-based molecules in computer switching devices.  Switches in the form of semiconductors are the heart of computers in semiconductors and storage devices. The state at a given moment is simply the sum of on-off states in storage devices made of semiconductors and information is stored on magnetic and optical media also as a series of 0's and 1's or otherwise put, "on" and "off" states.  The Holy Grail of computer technology is to find complex switches having a few basic characteristics.  1:  speed :  A device needs to switch, in other words change from the "on" to the "off" state at incredible speeds in the modern computer somewhere in the range of  billionths or trillionths of secondst .  2: Stability or reliability are critical in order to guarantee the states of the switch do not change unless we purposefully change them and the switches need to be durable enough to survive ordinary environmental hazards with no breakdowns. The switches must accurately record changes we make in them and preserve those changes until purposefully altered.  3: High storage density is critical.  Huge numbers of such switches need to be placed in a very small volume in order to make storage devices and processors useful for desktop and notebook computers.   Storage devices must be small.  Hopefully a light or laser will be able to alter the state of the device, in other words, to write a series of "on" or "off" state changes into memory also to be able to read written changes on memory devices.  In order to accomplish these goals, researchers seek to incorporate biological molecules into silicon computer devices.  Molecules such as Rotaxanes which are unusual large molecules whose structure may be altered by beams of light, and Rhodopsins molecules that are multiply altered by beams of light used in the eye and by organisms to store energy.  Silicon devices may be impregnated with biological molecules and beams of light used to "write" to these molecules, in other words to alter their state.  Some of these molecules change variably, depending on the color of laser light shown on their surface and seem to be very stable, holding an alteration of their state until changed by other beams of light. These molecules may then be placed in a three dimensional structure appropriately addressed for location so that a storage device is produced.[2]

In the field of Optimization, biological strategies of genetics and evolution apply to design of computer software.  You may wish to create a model strategy to improve financial return.  To do so you write a formula.  This formula achieves maximal financial gain under current financial conditions.  You can write a formula but you have no idea how it will do in the real world, a financial "habitat".  Why not take a hint from biology and use the methods of survival of the fittest.  Hold onto the financially lucrative parts of the formula and jettison the weakest concepts in a real financial milieu instead of trying to fly by the seat of your pants and create a theory that may or may not benefit you in the real world. This formula, is the most fit, that is, it achieves the greatest financial return.  A formula is either more or less fit than other formulas, which means that it either is a better or worse strategy for survival within a specific financial milieu.  In biology we mean by fitness the ability to pass down the largest number of viable offspring carrying our own genes, the ability to pass down one's own genetic endowment.  But this is a useful topic computationally as well and here is why.  Let us suppose the financial environment changes, stocks no longer are a good investment because of inflation or something of the like, then that was once the most adaptive fit formula now no longer is and another formula, will give us a better financial return. Chances are the second fitter formula shares a lot of the characteristics with the first formula, that it is related to the first formula almost genetically.  These optimization or fitness formulas have similar characteristics.  They are quasi-biological entities within a financial world and may be interpreted genetically.  As the financial environment changes these formulas for optimization of return need to adapt.   In order to do so they will have to change slightly, to mutate.  Or, possibly some of the inherent structure of this optimization formula will be borrowed from another formula in a form or translocation in much the same way as genetic material from one chromosome moves to another.   In the financial as well as the real world there is the survival of the fittest.  Many characteristics of these formulas may be borrowed, passed down and recombined in just the same way that genetic characters are.  Financial models may borrow from biology and vice versa. The biological method thus turns into a new means to seek truth and to pass it down.  More than this it gives us a design plan that adapts to changing habitats. 

What we are experiencing is the intrusion of biological concepts into computation and computational models into biology.  This is beside the point of attempts of computer scientists to achieve a level of parallel processing that is commonplace in the brain.  Though the brain is a Carbon based organic structure and computers are Silicon based, change is inscribed in both devices in a similar manner, electrically through transfers of charges and alterations of chemical molecules.   It is reasonable to expect that combination devices will be employed in computers incorporating carbon based biological molecules but also even more tantalizingly, silicon based devices may one day be implanted within the human brain.  These devices may aid in functions that are deficient in most humans such as computational and analytical skills.  If so human characteristics would be altered for good and there is the very real probability that the genetics of an individual, which is the Carbon-based living part of a person may not be the only characteristics that need to be preserved in future generations of progeny.  Such science fiction movies as "Total Recall" and "Johnny Mnemonic" have already begun to incorporate such concepts as Silicon based structures placed inside the cranium.[3]

If such fanciful mergings of living organisms and machines never come to be, carbon and Silicon are bound together anyway.  The computer revolution is unfolding right before our eyes and needs no amplification within these pages.  The computer will magnify human cognitive capacities in much the same way as the invention of writing or the printing press.  These inventions allowed us to record our thoughts and inventions to develop and communicate a collective consciousness, to build upon the past and work on complex concepts one painstaking part at a time.  Extemporaneous thoughts and music are primitive by comparison to recorded words, plans and music.  You might store in your own mind some sort of vague notion of the shape of an airplane.  But write these plans down, find a way to experiment and manipulate these plans, and work on them part by part with the input of experts in various fields necessary and you will design a real flying machine.  In music compare the primitive percussion of tribal music to the opera or the symphony. 

Today a single human brain can be connected to information in any corner of the world.  In his head is a certain picture an organization of his world, but inside his personal computer, is an alter-organization, a different world view which is also his.  With a minimum of effort and skill, an ordinary human brain can be connected to the total knowledge of the rest of the globe[4].  The major impact is expansion of consciousness.  We use an appliance outside the skull, not a piece of a biological organism inserted inside a computer, or of a Silicon instrument inside the skull, but some form of intimate contact between brain and machine with any of a variety of communication devices that may range from the traditional keyboard or mouse or touch screen or other pointing device to some kind of a virtual reality instrument.  This allows a person to point with his eyes, for example, or other body parts using thousands of tiny points within a virtual reality suit called 'tactiles'  (as opposed to visual pixels).  The purpose of this information appliance is to extend the abilities of the brain. At each contact the person would focus his attention or consciousness on the contents in the device.     

The concept of computers seen in science fiction novel is and on television is that of an advanced logical processor. Responses can be predicted as a function of hardwiring and soft­ware and computer logic is entirely deductive. By contrast, humans are more capable of inductive reasoning, able to recognize patterns. Humans can intuit and go from one topic to another fasciley considering a number of aspects of a prob­lem simultaneously rather than being tied to a sequential method. Reasoning is frequently done through analogies or may even hang from a thin thread of similarities or symbolism as often happens in dreams, myths and stories. Pure logic is only mode of mental operations. Mechanists fail to see the entire spectrum of human mental operations.

The effects of such nonlogical parallel mental processes sometimes surprise me as in a delirious or schizophrenic patient moved by thoughts that appear to be irrelevant or contradictory. Yet these abberencies show how thought is driven by a different engine than machine logic. Human thought may also be saltatory or jump­ing after the method by which electrical impulses are most speedily conducted in nerves. Long nerve cell extensions, the axons, are covered by myelin a fatty electrical insulator. Elec­trical charges cannot cross this insulated barrier. But at cer­tain intervals over the axon, the myelin insulation is interrupt­ed by discontinuities, the nodes of Ranvier. At these nodes elec­trical charges cross and collect. Changes in electrical currents are conveyed down the long axon by a process of jumping termed saltatory conduction. This method for the spread of currents is extremely efficient and fast. It is similar to the most creative human thoughts which don't rely on a continuous logical process but instead occur through discontinuities and the buildup of disparate motivational and informational factors just as charge builds up on an axon membrane, which can make a seemingly revolutionary thought almost inevitable. This phenomenon has been noted repeatedly and goes by many different names depending on the field of endeavor. In psychology and religion much is made of thought processes that are essentially foreign to a computer, designated as "aha" experiences or revelations, and are consid­ered to be bursts of understanding.

 


 

Figure 2: Neuron showing Nodes of Ranvier that allow electrical charge to jump thus speeding conduction.
 
 


[5]

 

 

 

The brain is an initiator of thought and feeling processes while a computer, at best, can bring an already initiated logical process to a successful conclusion. Even where it seems to be creating, for example, in providing the first move in a computer game, it is merely choosing from predetermined alternatives. If we know everything about  computer hardware and software and we can define the stimulus, then we can always predict a response. Some neurobiologists believe the same about the brain. They look at the brain rather rigidly as a hard-wired complex of conduction pathways and circuits. Upon this is superimposed human experi­ence, perhaps learned patterns of response analogous to computer software. Know everything there is to know about how the brain is built and functions and also its experiences and you will always be able to predict accurately, its response. There is a certain smugness about this mechanistic all-knowing approach, also a certain amount of backward reasoning, an assumption that computer circuits simulate the brain's function. It's much easier than admitting that human thought and action is not determined or at least that we have inadequate data to have an opinion about whether or not it's determined. To the biological mechanists human thought, feeling, and action would be one hundred per cent predictable, if only we knew more. Our inadequacy in prediction comes purely from a lack of knowledge. The observed randomness of behavior is only an illusion or false perception following from our ignorance.

Early in the twentieth century when physiologists had finally described the simplest of all nervous system responses, the deep tendon reflex, they naturally became infatuated with the idea, and an artificially inflated notion about the level of their own understanding. The brain and nervous systems responses, they reasoned, must function  only as a complex of simple reflexes. If we knew everything about all reflexes then all of the brain's responses could be understood.  As it turns out, higher nervous centers serve mostly to dampen the stretch reflex. The higher order neurons of the pyramidal tract synapse directly with spinal cord motoneurons, not only to convey commands from the brain, but reduce muscle tightness or tone. Others are connected to control of muscles changing the stretch receptors themselves (termed the gamma efferent system). This brings up a basic principle of nervous system function. Higher order neurons that control simpler lower order circuits do so mostly through modulation of the hard-wired lower order reflex response, in other word through inhibition.  These physiologic understandings are expressed clinically in various neurological conditions. When the upper part of the spinal cord is interrupted, the stretch reflex in the lower part of the cord controlling the legs, is liberated from inhibition of the brain.  The stretch reflex in the legs functions with impunity. The afflicted individual has a very active deep tendon reflexes and very increased muscle tone, termed spasticity. Many people think they are healthy if their reflexes are active. Actually, the opposite is true. Muscles become very tight even when moved passively by an examin­er suddenly giving way in a clasp-knife manner through their excursion around a joint. The spastic will experience at best a very tight scissors gait and at worst, even at rest his legs will tighten or bounce continuously as his uninhibited stretch reflexes express themselves.

Drugs may be used to help such patients decrease muscle tone. These may inhibit motoneurons or affect gamma efferents or sometimes may impair the mechanism of muscle contraction itself. Neural circuits are deceptively simple and predictable only when studied in isolation.

Hard-wiring neural circuit­s are in higher animals and even man means a predictable  series of electrical responses.   We see this in a record of tiny electrical potentials in the auditory pathway of the brainstem. In a test called the Brainstem Auditory Evoked Response an auditory stimu­lus of short duration (a click) will reliably produce a series of 5 electrical bumps or waves as the nervous system responds by conveying the message from the lower brainstem to higher centers. This series of waves will always occur, each bump representing a way station or synapse that corresponds to an anatomical point in the auditory pathway (Figure 4). This is most useful as a test in medicine. If a bump is absent or delayed this is a sign of a problem in the specific area corresponding to the wave. After the impulses travel through this well-established pathway, electrical responses become much less predictable. This corresponds to the role of higher nervous centers (the cortex) where sounds registers in consciousness.   This is a pivotal point that should not be lost.  Another thing about nervous function at higher and higher levels is that responses are less stereotyped and predictable.  The higher one gets within the nervous system, the less predictable the response. As a sensory impulse travels through the nervous system over the first few synapses the pathway is determined.  However when you get to the cortex the electrical response is widely distributed and cannot be predicted at all.  More primitive organisms have only this stereotypical nervous response.  It is a sign of more advanced nervous function that you stop being able to predict a response one it achieves a certain level. It seems to me this is a general comment on higher and lower function in general. 

Computer scientists are becoming less naive about nervous function as they discover that machines just cannot duplicate nervous function or reason even at a child's level. There has been some apprecia­tion of the function of networks of neurons and function of such units. Each individual nerve cell is literally connected to thousands of others. A neuron's output may directly connect with hundreds of others through axonal branching and dendritic spines provide a much more complex array of inputs. The decision whether or not to fire may be influenced by many thousands of nearly simultaneous excitatory and inhibitory inputs from other neurons. In many instances a single neuron’s output may return to act as a feedback mechanism inhibiting further firing. Scientists are working on ways to monitor the output of arrays of neurons rather than single cells. Cultured single layers of neurons have been monitored using arrays of tiny electrodes. With this apparatus it is possible to "listen" to the integrated output of groups of neurons. Some information has already been obtained showing patterned firing in groups of cells in these primitive cultured networks. Though this does not reproduce by any stretch of the imagination, true brain function, it is an attempt to get at the sociology of neuronal response.

Figure 3: It is an important principle of nervous system function that as a general rule, the more advanced the nervous system response, the less predictable and stereotyped it is.  The more synapses the less you can predict the response.

Figure 4: The auditory evoked response.  Each electrical potential corresponds to an anatomic waystation in the auditory pathway. Each wave is a reliable stereotyped electrical reflex beneath the level of conscious awareness[6].

 

As we have seen the idealization of brain function in terms of sequences of on-off responses is suboptimal. There have been attempts to set up arrays of circuits which may simultaneously affect each other's output as neurons do, which is more like how neurons in the wild state interact.  Higher thought sometimes involves the combined effects of numer­ous stimuli with which we have to make due even with incomplete information. Any child can recognize a person from his voice, facial characteristics even given such constraints as inadequate light, camouflage, etc. A manager has to decide on an optimal strategy for completion of a task, simultaneously considering different elements of that task, for example, differing abilities of various staff members. This is fundamentally different than strategy in a chess game, which occurs one move at a time each move affecting only subsequent advantage. Such arrays of cir­cuits, each of which may simultaneously affect the output of other circuits, more precisely mimics actual brain function. As of yet such arrays function only on a primitive level, but they more faithfully reproduce brain processes that are usually nonse­quential.

The nervous system is not static. We know that synapses constantly break down and form anew. Each time a nerve cell dies, thousands of synaptic connections are destroyed. Learning also reflects in anatomy as new synapses, connections, form.  You lose mental capacity with age, Alzheimer disease through the misuse of drugs and degeneration you are really destroying more and more synaptic connections between neurons.  Thus although you generally lose neurons with time and with them abundant synapses, you form others through learning and mental exercise.  If you continue to learn as you age by doing problems, reading, and expanding your vocabulary, you will form new synapses.  Adult life is a race between neuronal loss and synapse formation.  It may be the total number of synapses formed that determines the net change in mental capacity.    This is exactly the same as preservation and increase in muscle mass and bone density with physical exercise which can also retards aging and preserves or even increases capacities.  Hence it may be that with mental exercise Alzheimer and related CNS degenerative diseases are staved off!!

After a nervous system injury even one producing large-scale damage and nerve cell death, functions that are lost mysteriously reappear.  It helps to foster healing through physical and cognitive therapies.    This is the macroscopic picture.  The real change takes place on the level of the individual nerve cell.  If the nervous system functions as an executive for other organ systems of the body, the neuron performs this function in miniature. On the grand scale sensory data must be organized, perhaps mulled over somewhat, but then acted upon. The neuron must also organize its response to complex inputs. Thus it may be viewed as a single molecular unit of nervous system function.

There are anywhere from ten billion to one trillion neurons in man. This estimate depends upon an accurate count of nerve cells (which is hard to come by) then the assumption that the density of neurons is everywhere about the same* (which it is not) and that this number can be integrated over the entire volume of the nervous system (which varies). The number of neurons in the brain is similar to the number of stars in a galaxy. Neurons and stars are energetic systems. Stars exert their affect upon their fellows gravitationally mostly over long distances and occasionally (in supernovas) with the explosive transfer of matter from one star to another. Neurons make more intimate contact with each other and also have their own entourage of glia and other supporting cells. The central nervous system, containing the great majority of neurons, is somewhat insulated from other organ systems by an advanced and highly organized blood brain barrier. Yet if there is a reason for the biological existence of man and all his physiology, it lies within the nervous system.          

The human nervous system controls or influences all bodily function. Quite a lot has been made in popular media, by authors such as Bernie Siegel and Deepak Chopra of nervous influence on immunity.  The thought has been that by influencing psychology (hence, indirectly brain function) a person's immune system may be made to play a greater role in diseases that depend on immuni­ty, for example to help destroy certain malignant tumor cells. The logic here may be far-fetched, but the general idea that the brain is the major locus of organic control of all such functions is not far fetched at all.  Nervous function arose in evolution alongside other organs sys­tems performing their own specialized functions. The nervous system seems to us to be physiologically all important, but it very likely arose as an afterthought not the primary goal of biological process at all, an epiphenomenon, in other words. 

Nervous tissue arose from the need in primitive animals composed of small groups of  cells for these cells to somehow communicate.   As soon as you have more than a few cells in an organism and specialization among them, you create the need for information transfer.  The development and use of nervous tissue as an arbiter of information transfer between cells is but one of many strategies for survival.  Brains and neurons have achieved prominence only in a small branch of the whole biological tree, namely among certain vertebrates and mammals particularly and among these especially in primates that have evolved only recently from the timeframe of biological evolution.   Animals employ different strategies for adaptation.  The vast majority adapt to their environment by changing their own biology over many generations, in other words genetically.  Some few animals, mostly advanced animals, are able to accelerate their adaptation and change along with their environment by learning.  In so doing, they may alter their response within a single or within few generations - a useful trick for organisms having a longer lifespan or long generation time.   Mammals rely to a much larger extent on ad­vanced nervous structures and have a knack for adapting over the lifespan of an individual animal.  Even in the biological sense then, an individual assumes much more importance.   By contrast, the most recently evolved invertebrates, highly successful ones  from the standpoint of numbers, competition and speciation, especially insects, de-emphasize  plasticity and learning.  Thus we have two entirely different formulas for competition.  Both obviously work.   Insects' nervous system responses are reflexive, stereotyped and dependable.  Though complex behaviors can certainly occur, these are hard-wired responses, genetically endowed.   It is no big deal that adaptation can occur only over generations.  The insect’s generation time is short,  the number if individuals extremely high, the rate of adaptation and differentiation fast.   Among insects the individual is de-emphasized.  In fact, for many insects individuals are very nearly genetically identical, true in particular for social insects.   The individual nerve cells of insects look and function very much like ours, but the system allows for little plasticity or learning.  Mammals and insects are in hot competition for global hege­mony. The farmer's constant struggle against insect pests illus­trates this.   As insecticides and other tech­niques are used insects speedily adapt, changing genetic features facilely over many generations. It's our brains and planning against insects phenomenal ability to adapt, that is, survive all of our attempts to reduce their numbers.

Bacteria and other microorganisms also show how living things can compete by genetic design.   Biology pits this genetics against human brainpower and no one can guess which strategy will prevail.   In the race for survival, the human brain doesn’t always win out.  Take a look at the mess we've made with antibiotics.  There must be well over one hundred of them in common use just in the United States.  Bacteria have an uncanny ability to mutate into resistance.   It’s getting so some strains have to be treated with two or three antibiotics at once, especially hospital-acquired infections, because these bacteria living in hospitals descend from strains that were exposed to and survived antibiotics.  The most dangerous infection to get is a so-called nosocomial or hospital acquired infection because these bacteria have already seen and are resistant to commonly used antibiotics.  For bacteria living in hospitals, their ecosystem contains our best and most potent bactericides and they have adapted to survive our most potent drugs.   Bacteria commonly pass down enzymes that deactivate antibiotics or the bacteria themselves may be infected with rings of genetic material (plasmids) that code for these enzymes and allow for survival in an antibiotic laded ecosystem.  Even taking into consideration the community outside the hospital, antibiotic exposure is rampant as patients demand to be treated for infections they don't have, minimal sinus complaints treated as sinus infections, discomforts in the urinary tract erroneously treated as urinary tract infections and so forth.   Organisms that survive antibiotics especially fungi such as Candida then have the advantage as competing bacteria are killed off and we then see minor complaints such as vaginal discharges and itching being over treated with fungicides and so the problem propagates.  Nursery schools are filled with toddlers all with middle ear infections all on chronic an recurrent courses of antibiotics that serve merely to foster reinfection and antibiotic resistance.   Our profligate use of antibiotics mirrors overused of pesticides and defoliants so that it can be said without exaggeration,  our very worst  enemy is ourselves with overzealous over used of drugs and misuse of chemicals.

Sometimes the bacteria win out. We’re having a terrible time with multiple drug resistant tuberculosis even though we’ve developed six or seven antibiotics (some quite toxic) in common use. These have arisen from incomplete treatment of a relatively few cases but the multiply resistant TB Bacillus once acquired, is almost impossible to eradicate.  TB is one of those organisms that lives within host cells for years and so is hard to get at with antibiotics or with our own immune surveillance mechanisms.  Other diseases of this type include Herpes viruses, leprosy and most importantly Malaria which still kills well over one million people a year and infects 300 million persons.  Bacteria and viruses that have an intracellular existence are often dormant and assymptomatic for years until they declare themselves with a chronic recrudescent infection or a fulminent acute clinical attack as does malaria.

The neuron responds to information converging upon it from one thousand or more other cells through up to ten thou­sand or even more synaptic connections. Neurons have their own functions, but have needs and  proper­ties at once similar and yet different form other cells. The most basic thing you can say about the neuron is that it is excitable. Thousands of converging inputs add up to a resultant firing or non-firing state and the ultimate decision rests with the neuron.  Whether the cell will or will not fire is the question determined not only by input, also characteristics of the cell.   Admittedly, as we shall see, some neurons particularly those involved in interpretations of sensations in the periphery, have a graded response, but most central nervous system neurons either respond with an action potential or they don't, giving rise to the expression, an "all or nothing",  response. What is involved in this decision?

 

Inside the Neuron: How the Neuron Works

A neuron will fire under certain limited conditions. That is properly what distinguishes it from other cells. Like other cells, it is surrounded by a cell membrane that delimits it from its environment. This bilipid (fatty) layer separates two watery environments, the inside (protoplasm) and the out­side of the cell. The membrane must discriminate between various substances it will let pass. Consequently, there are differences in concentrations of various substances inside vs. outside the cell. Sodium is roughly fifteen times more concentrated outside the cell, but a similar but larger singly positive charged ion, potassium, is nearly thirty times more concentrated inside. A chemical pump moves these ions maintaining these critical concentrations against electrical potentials and a concentration gradient.   Because there is a tendency for sub­stances to diffuse from areas of high to lower concentration, force (and hence, energy) is required to maintain any concentra­tion gradient on the two sides of the membrane.

The sodium-potassium pump gets this energy from the nearly universal final common pathway for cell's immediately accessible energy, the molecule, adenosine triphosphate (ATP). ATP is pro­duced out of the energy supplied by other chemical compounds, especially for the case of the brain, out of glucose. This chemi­cal energy conversion is the job of the intracellular mitochon­dria. These independent organelles are fascinating because they contain their own genes unlike no other animal cell organelle. Plant chloroplasts actually have a very similar structure and do the opposite of mitochondria. In plants chloroplasts fix energy absorbed from sunlight into chemical energy storage in the form of sugars. The mitochondria's job is to extract the latent energy from compounds especially sugars.

Mitochondria and chloroplasts are membrane bound and have an advanced lamellated structure, almost as if they are organisms unto themselves. They are the only organelles that contain their own DNA and they reproduce independently. Mitochondria have almost a symbiotic relationship with the cell. A lot of mitochondrial protein is coded for in the nucleus of the cell, produced in the cell cytoplasm then imported into through the mitochon­drial membrane. These proteins are attached to a special series of amino acids, in other words a short peptide, that acts as a code which tells the mito membrane that the protein is meant especially for the mitochondrion and needs to be imported. After importing these critical proteins, the mitochondrion uses other special proteins that cleave it, leaving its active portion. But certain other critical proteins are encoded by special mitochon­drial DNA. There are a few interesting diseases that are inherit­ed problems relating especially to mitochondria.

These disorders tend to affect muscle including the heart, preferentially. This stands to reason when you consider that mitos deal with energy utilization and muscle’s tissue is the most energy intensive tissue in the body. Second these disorders are passed down only through the mother. Your mitochondria are passed down through the ova alone and are not inherited through the sperm which carries only DNA from the nucleus. The mitos reside in the cytoplasm of the ova. This shows that the sperm merely is a vehicle that carries information from the nucleus of the fa­ther. On the other hand the mother's ovum does carry many geneti­cally divergent mitochondria. In some of these disorders mito­chondria are extremely plentiful. The cells especially muscle cells can be full of them. Maybe there is a feedback mechanism that tells mitos to reproduce if their job isn't getting done. Alternately, they may appear to be wildly abnormal under the microscope. As you might expect, they don't work well either. But there are found to be specific problems when we look for them, ordinarily with coding for a single protein as in other genetic diseases. Persons with these diseases may have something wrong with their heart rhythms, may have muscle weakness or fatigue or may not be able to process some energy containing foods especial­ly fats and so become weak and fatigued on that basis.  Just one typical example is a disease that affects only the transport of certain fats into the mitochondrion so energy can be derived. This in­volves a substance that helps to transport these fats called Carnitine. The brain usually functions abnormally as well.  Mental retardation is a problem and the tiny muscles that move the eye will often be affected so that eye movements can be grossly abnormal and a person may have double vision. The diseases are often progressive and severe. Under the microscope a typical abnormality is what is called a ragged red fiber with grossly abnormal and proliferated mitochondria. (Figure).  But the moth­er's ovum passes down some of these diseases. Others are not passed down probably because they are so devastating that a person with the disease does not typically reproduce. We think that there is an alteration in the Mitochondrial DNA early in life (a mutation) and that the disease occurs sporadically (non- genetically) on that basis.

Mitochondria and chloroplasts may be remnants of primordial bacteria or other small organisms that continue to live in a symbiotic relationship in every animal and plant cell. Many of the cell's organelles seem almost to be organisms unto themselves. Another example of a structure in the cell recently discovered is the peroxisome. These structures went unnoticed for many years, but are plentiful in kidney cells so that after a certain period they could not be ignored. They get their name from certain enzymes they were found to contain that help process peroxides and help to detoxify and protect the cell against free radicals that can harm it. Later these structures were also found to be important in the metabolism of fats especially long chain fatty acids and disorders in these structures among other things cause diseases of fatty acid metabolism the adrenoleukodystrophies. These diseases are often present in males predominantly (are x-linked) and cause adrenal gland insufficiency and affect the deep white matter of the brain where these fats reside. Peroxisomes may one day prove important in detoxifying free radicals which are implicated as the cause of neuron destruction In many chronic diseases especially brain diseases.   But the point is the Peroxisomes are similar to mitos in some ways being delimited by membranes and almost separate from the cells containing their own enzymes made in the cell but targeted for these structures specifically, most probably by a mechanism that is similar to the mitochondria described above. Mitos though are so basic to biological organisms. All of them have to provide energy for the business of life and so mitos are present in every cell and even in the most primitive cells.  Cells so primitive that they do not have a separate nucleus, still have mitochondria without a membrane delimited nucleus ie bacteria, proving that they must have become intrinsically important to life, very early in evolu­tion. If they come from a certain symbiosis, it took place ex­tremely early in geological history and is extraordinarily basic. In fact both mitochondria and Peroxisomes and chloroplasts too are thought to have evolved from an early bacterial or blue green algae invader of ancestral animals that eventually became a symbiote, its DNA encoded now in the nucleus of the host cell.  Peroxisomes may have been in the cell earlier# and were useful once because oxygen was lethal to the primitive anaerobic cells that were at the beginning of evolution as oxygen is even today to anaerobic bacteria such as Clostridia.  As time passed the peroxisome took on other chemical roles such as handling l