4.2.1 The Debate Over the Brain

It is said that the hand that rocks the cradle rules the world. Perhaps not, but it can contribute small victories in humanity's long struggle to understand the human brain. The February 3^rd 1997 issue of Time showed an intent but happy baby on its cover. Inside were more pictures of babies being nurtured and parental advice on what was required, including tips from the First Lady. The article spoke of the 'wiring' of the brain, but the remainder of the message was clear. The human brain was not a computer that was assembled and then turned on. It was a living, growing, nurtured device, with environmental influences beginning in the womb. The article explained that genes wire the basic motor circuits of metabolism and encode general instructions on how to build a brain. But most of the infant brain, even for motor responses like vision, is learned, and learning effects were startling. Children raised in non-stimulating environments developed brains between 20% to 30% less than normal size. Even rats raised in stimulating environments developed up to 25% more synapses per neuron than ones in drab surroundings.

To people familiar with the unfolding debate over the brain, none of this was a surprise. Three centuries early John Locke proposed that the brain was a tabula rasa, or blank paper onto which the experiences of life were learned. This century testing by Pavlov, Skinner, and the child psychologists Jean Piaget had established the basic learning mechanisms of the brain, including those of child psychology. Even one startling revelation of the article, that the infant brain is effectively rewired between birth and three years had been known at least a quarter of a century. Other known facts which could have been illustrative, such that at birth a human child has only 25% of its adult brain size, compared with a chimp's 65% were not in the article. But the crucial arithmetic of why it is impossible for the genes to encode more than a fraction of the brain's total knowledge was given public airing. In the arithmetic of the article "quadrillions" of connections are required out of some 50,000 genes available to specify how the brain is wired, so the rest must be a learning process, even if one whose basic outline is in the genes. With this can be thankful for small victories, and that parents of the world wanted from science practical knowledge about raising children, not ideology. Although the science of the learning and nurturing of the brain was already known in the public mind we seemed prepared to abandon one of the more ridiculous scientific myths of this century; the alleged wiring of the human brain.

"The newborn infant is now seen to be wired with awesome precision… " writes Wilson in his book On Human Nature. He wonders to what extent this wiring "so undeniably encoded in the genes" will determine all development. Still, in many ways the metaphor for wiring in the brain is often unavoidable, even in this book. The everyday term for complex devices controlled by electrical signals is 'wired'. And brains are just that, with many connections looking remarkably analogous to the wiring humans use. Only while the metaphor might be unavoidable, the thesis that genes determine human behavior because they precisely wire each neural circuit is not. Wilson posed hard questions about humanity but he never quantified to which extent the "wiring of the neurons" effects behavior, and wiring remains a misunderstood metaphor. So beyond the metaphor, to what extent is the brain wired really?

Firstly, while the metaphor refers to wiring, axons and dendrites of the brain form tubes rather than wires. These tubes carry an electrochemical fluid along the tube canal. The distinction is important. Electrical wiring is firmly anchored into a terminal, whereas the tubing of the neural circuits has variable connections but few of which are solid. Many connections can be altered or reconnected during the neuron's life, but the most dramatic changes are during growth spurts of infancy, especially the birth to three-year-old period. This is why parent need to know. An electrical wire will not flow current until it is anchored both ends, but the tubing of the brain can flow an electrochemical bubble of potential along it similar to how water can flow down an unsecured hose. This is why stimulating the newly formed brain is important. Research has found that flow of electrochemical potential along the tube stimulates brain growth. Roughly, tubes that get good use find suitable connections while those that atrophy from lack of stimulation suffer retarded development. This starts in the womb, and experiences a great spurt after birth, though parents need to consult experts about the sequence.

However, tubing of the brain allows not only learning of basic design but a subtle process where the tubing connects to other axons or dendrites, at a junction known as a synapse. Unless it is a bad joint, an electrical wire is either connected or it is not. But synapse joints have a degree of "memory" due to the electrochemical nature of the join. If two copper wires were solidly joined to a terminal and one wire flowed current thousands of times a day and the other did not, the joints would not remember apart from getting hot, which one was more used. But by electrochemical modification the synapse can remember how many times it fired in the past, which affects its fire rate in future. If we call the solid tubing of the axon or dendrite paths "hard" learning, modification of the synapses is "soft" learning, known as synaptic facilitation. Thus we have not one neurological learning process in the brain, but two, both of which allow extra-genetic modification, and neither of which bear an analogy to the "hard" copper wiring of a switchboard or telephone exchange.

The third break of analogy between switchboard wiring and the brain is the firing process of the neurons. In a typical circuit, many dendrites feed a neuron cell (known as a soma), plus axons from other neurons feed the cell via synapses. When the soma 'fires' it sends a signal down its own axon. But the actual firing of the soma results from not one signal, but many, and there is a potential build up from many synapses. The neuron makes a weighted decision depending on how many synapses are ready to fire, other signals from the cell's dendrites, presence of electrochemical stimulators or inhibitors as moods, and past learning experience of the soma itself. Probability effects of single electrons could also cause firing, especially on finely balanced potentials. So, each neuron has a tiny amount of "free will" in its decision to fire, which might only delay a decision to fire by milliseconds for a single neuron. But the human brain contains 100 billion neurons all interconnected with each other. It would not happen this way, but for orders of magnitude a one millisecond delay for ten billion neurons summates to three years total delay.

This random variation to how each neuron fires is not an explanation of will. But knowing of mechanisms like this in the brain, we should not rule out random behavior because of analogies to how a telephone switchboard is wired, knowing that a telephone switchboard does not posses mechanisms remotely analogous to neuron firing in the brain.

But not only is the brain's analogy to electrical wiring dubious, strangely, so is the reverse analogy. If the brain is optimally designed biologically, and we applied optimal design methods to circuits used for control engineering what would we end up with? It would still not be "wiring" of the type people like E O Wilson seem to think is in the brain. This can be shown very simply.

Firstly, at any stage of design of control circuits, there must exist a differentiation of what can be called hardwiring and softwiring. At a basic level it is just that; hardwiring would be a tougher, less flexible wire, softwiring would be less rugged wire, but more easy to bend or thread. At another level softwiring would be programming circuits into a purpose built electronic controller, while hardwiring would be physical wiring. At another level, hardwiring would be the purpose built controller, softwiring would be a general-purpose computer program, and so on. For any control system, the rule would be; "hardwire safety, softwire process". So how does this work?

Consider a simple water pumping system. For any pump there are certain safety conditions which apply regardless. For instance, one would not run a pump dry of water as it would destroy the pump. And for any large mechanical device there would be a hardwired lockout switch to immediately disconnect power in an accident, and so on. These minimal safety devices would be hardwired into the control circuit of the pumps so the pumps could not be run without them. However, the process is different and could alter in ways not affecting safety. You might run alternate pumps on alternate days, or run more pumps on weekdays, less on Sunday, and so on. Whatever it is you must allow for change. So providing all the safety devices on the pumps had been hardwired in you would softwire the process to make it easier to change. Just you would know that no matter how much the process control was changed, even for a mistake, the hardwired safety circuits would apply regardless. It is the same with traffic lights. The process of changing traffic lights can be very complex, but no matter how much traffic light signals varied the circuits must never allow any lights to fail green, such that lights facing intersecting traffic could all be green at once. Therefore, even for computer controlled traffic light system there should be hardwiring such that even if the computer instructed all the lights to turn green the hardwiring would still prevent such an instruction being carried out.

This is why we should not claim that the brain of a newborn infant is "hardwired" without explaining which circuits are hardwired, or why nature hardwires some circuits but softwire others. As explained in the previous chapter the embryonic hindbrain is totally wired at birth, with only the more recently expanded cerebellum partially wired shortly after birth. On the other hand the forebrain is mostly loosely connected at birth and wired postnatal by the learning process. Nature takes these basic building blocks of brain types and selects a combination of neural function appropriate to the survival needs of the organism. Humans, whom Wilson discusses, have roughly 80% of neural mass in the higher, learning cortex, about 12% in the slightly modifiable cerebellum, but only 8% in the remaining parts of the lower brain hardwired totally at birth. Part of the cortex is hardwired, but humans only having 25% of adult brain mass at birth compared to the chimp's 65%, or a twenty fold increase in synapse connections after birth indicate how learning intensive human brain development is. In primitive organisms most neural functions will be hardwired because for primitive organisms most response to stimulus will be reflexive. Only hardwired circuits run into the problem of genetic transmission of total bits of information that can be encoded in genes to the total bits advanced organisms need in their brain. From the previous graph we see that simple mammals require about 100 billion bits of information while humans might require upwards of a trillion bits of information. This means that humans require an order of magnitude more bits of information in their brains than alleged wiring of the neurons could encode.

Even so, there must have been an optimal ratio of learned to reflexive design, which we can call it the learning ratio. If engineering were a guide the reflex circuits would be safety and hardwired, while learning circuits would be process and softwired. This means that the greater the ratio of learning circuits the less instinctual hardwired safety circuits any creature would start life with. Unfortunately, there is no data on what this ratio is so we can only infer learning circuits occupy the higher cortex, while reflex circuits occupy the more primitive structures of the brain. All vertebrates have at least some forebrain, which appears to contain primitive learning from the start. However, as life evolves in complexity all segments of the brain grow but the forebrain grows fastest. Again, in primitive vertebrates the forebrain, midbrain and hindbrain each take up about a third of the total cranial capacity. But in humans while all segments of the brain grow, the forebrain grows seven to eight times larger than the other segments combined.

Humans also have about three times the brain volume of a chimpanzee (1350cc for man to 450cc for a chimp) but about four times the area of higher cortex. This gives a two-to-one ratio of learned to reflex neurology for a chimp. Yet the ratio increases to about eight to one for a human brain. Assuming roughly, chimps and humans posses the same number of reflex circuits in proportion to body weight. More dramatic is the much larger frontal cortex in humans, about 29% of the human higher cortex against 17% of the chimp higher cortex. It seems that if the normal cortex results from evolutionary encephalization of functions previously used by reflex the frontal cortex is a further encephalization of the learning and planning functions of the middle cortex. This gives an approximate frontal cortex to reflex ratio of 40% in chimps, but about 230% in humans, or the human frontal cortex about 7.6 times the size of the chimp frontal cortex (see Table).

Moreover, while absolute brain size is a function of body size for reflex, for the frontal cortex free from the needs of reflex size will be in direct proportion to capability. In these terms the human brain could posses almost a 99% commonality of circuit design with a chimpanzee, but still be a radically different brain because of the ratio of reflexive, to learning, to prefrontal cortex circuits. We see the evolutionary design advantages of quickly expanding the neural capacity the human brain this way. It probably took as much evolutionary design for the large human cranium to egress the female pelvis, than to multiply the learning neural circuits this way.

But why is the learning to reflex ratio so high in humans?

Clearly, it is an easy way to expand brain and learning capacity. If the human brain needed to be large and available genetic transmission was saturated, learning circuits are the only way to expand brain size further. But, an excessive learning ratio creates dangers. It makes the human infant exceptionally vulnerable, and requires long periods to nurture and raise the young. It also makes a young human very group dependent because virtually all its survival skills must be learned. Plus the large cranium vastly complicates the birth process making unaided birth impossible for the human female. So there must have been great evolutionary advantage to this high learning ratio of the human brain, or nature would not have pushed it to such biological limits. So, what were these?

We can only return to the premise the human evolution was filling out the last environmental niche on Earth, by a process of general adaptation to all possible environments. The process once set in motion did not stop until complete optimization of the biology was achieved. Earlier chapters explained why a brain capable of imagination, reasoning, and abstraction provides maximum options and the very high learning ratio achieves that design end. But why this particular ratio? Or more likely, why did the brain keep expanding until some apparently optimal ratio? The optimal ratio was determined by natural selection, so we must ask why say, at 7.5:1 the evolution of the brain was not complete, but at a slightly higher 8:1 ratio no further modification was required?

This is the crux of the modern argument. The qualities the human brain best displays, judgment, intuition, imagination and reasoning require a breaking of the chain of cause-and-effect events throughout the universe. The brain needs to do what humans often need to do socially, isolate themselves from active surroundings and think things through. There are nine billion neurons in the human brain, more than there are stars in the universe. Some 90% of these will be associated with the learning process, which cannot be exactly mapped to genetic cause-and-effect. Especially, there will be a small random noise effects of the electron firings at each synapse. So, we presume the inner neural circuits among the eight billion odd learning neurons of the human higher cortex will have a degree of pure reflection to their operation. The high learning ratio of the human brain might be an evolutionary critical mass of the minimal required number of modifiable neural circuits, to achieve abstraction, reflection, and contemplation in the brain. Of course whales and dolphins have large brains too, only this might not mean that whales can abstract. The critical requirement seems to be the learning ratio itself, and the amount of pre-frontal cortex, but there is little data on what this should be.

This very high learning to reflex ratio in the human brain coupled with the brain's unique abstraction qualities, which are the likely a consequence of it, allows us one more parting remark on the wiring of the brain. When Hume debated whether the features of the world might liken it to a well-designed house, he argued that from a viewpoint of intentional design, you could as easily liken the world to a vegetable as to a house. Similarly, if asked what human artifact the brain might most be likened to, a good answer might be a thermos flask! Just as a thermos flask isolates heat information from the inside to the outside of the flask, so the cerebral cortex of the human brain uniquely isolates the cause-and-effect information outside of the brain from that which is allowed to occur within it. And this is exactly how humans come to use imagination, reflection, and symbolism to abstract from the physical cause and effect of nature, what their real options are.

This thermos flask analogy of the human brain, proffered half-jokingly, would if even partially correct make the human brain a very unique device in the universe; certainly a device we have no idea how to build with present technology. We always suppose that though no equations describe it, every action in the universe mechanically influences every other. And brains are clearly part of the universe responding to its inputs, that is their function. Only to return to the troubling dualism of Descartes we wonder how there can exist in the brain thoughts that are perfectible, such as mathematics, in a universe that is not perfectible. But we never get the perfectible thoughts of mathematics to align completely with physical reality. Mathematics and logic remain a category of knowledge we call analytical truths, while the physical events of the universe, which we measure, we call empirical truths. Plus there is this disputed category of moral truths, which do not exactly fit as analytical or empirical truths either. This separation of categories is necessary to maximize the options of knowledge. It allows us to apply analytical truths as proof of non-contradiction to any range of problems we care to conjure knowing these truths are not tied to physical categories anyway. Plus it also forces us to test our assumptions as analytical truths, against the measured properties of the physical universe as it exists. This separation between what exists and what we can conjure forms in the higher cortex of our brains.

Even so, this separation in the brain between perfectible thoughts, which might exist slightly disconnected from physical input, seems to be the whole issue. It is not that theories of hardwiring the brain are scientifically futile in explaining how the brain functions, it is why people persist with such analogies at all. In the long history of debates over the brain the use of mechanical analogies to describe its function, whether it is springs, levers, clockwork, mechanical motion, or wiring, or even thermos flasks, is no surprise. It is not whether the analogy is valid, it is the impression that the purveyors of these analogies try to perpetuate. This is that the universe is a gigantic clockwork whose parts can be disassembled, and which functions by a rigid train of mechanical cause-and-effect "set in motion before we were born". This causation mechanism supposedly dispenses with the intervention of God in the workings of the universe, so we are able to describe cause-and-effect in secular, scientific terms. Only the strict mechanical cause-and-effect seems to preclude, in some people's view, any moral debate over human purpose, and this is the big issue. Wilson say, who praises homosexuals as "genetic carriers of some of mankind's rare altruistic impulses" also regards cloning or conventional eugenics as a way to improve mankind. These are all controversial moral issues that we will not be stopped from debating on grounds that the brain already locks us into mechanical causation, which we have no control over.

Presenting the brain as a mere algorithm forces critics into an 'either/or' position on how we view evolution. We must accept evolution as interpreted by ultra-Darwinists such as Dennett, Dawkins or Wilson, or be accused of philosophical faintheartedness or even secret religious sympathies! We are not say, allowed to question if there might not be emergent effects from the large neural brain bulk, or debate the cultural impacts of such a large amount of post-natal learning in the human brain. And we must accept this ultra-Darwinist view unquestioned, even when its advocates get themselves into problems over how to explain the brain. Dennett say, has no explanation of how the brain works culturally other than by appealing to the discredited idea of memes. Memes are central to Dennett's thesis, yet Wilson never mentions them, but goes of on his own tangent about eugenics and the alleged altruism of homosexual genes. Other ultra-Darwinists are off on the peacock's tail effect, or humans having large brains because they like to gossip. All these theories have serious flaws to an extent of misstating, often grossly, the basic neurology of how the brain works.

But does science really teach that the universe operates by a rigid chain of cause and affect anyway? Do humans as a species uncover scientific laws merely to become slaves to them? Or is this just one more hardwiring of the brain myth, not a scientific law itself but a mere expression of an ideology?

Let us now examine the problem of cause-and-effect in the human brain.