Rocket Science & Human Behavior

Placide D. Nicaise

This is a discussion about human behavior, but before I can explain my point of view, I need to discuss a process that is more than an analogy between mechanical and biological systems. It is exactly the same process, employing the same elements, and meeting the same objectives in both machines and animals. This process, known as feedback control, is fundamental to robotics and machine intelligence, just as it is fundamental to all living things. It is no accident that man adopted the use of feedback control as soon as machine technology permitted it. This is the process that turns a "dumb" machine into a "smart" machine. It gives both machine and living things the uncanny ability to relentlessly pursue preset goals. In one case, a human programmer sets the goals, and in the other, nature sets the goals.

I don’t intend to make this discussion of feedback control mechanisms seem like more than it really is. It is not magical. This is not another fantasy about machines coming to dominate society. Feedback control is a well-known, highly refined and specialized technology that has been widely used for decades. The engineering details of feedback control are not necessary for this discussion. It is only necessary to understand why and how it works in machines, so it will be obvious that it is the same process at work in animals.

A feedback control mechanism has three elements: (1) sensing, (2) computing, and (3) actuation. The sensed parameter can be almost anything, such as temperature, pressure, angle, position or speed. The computing element must contain some preset target value. The sensed value is "fed back" to the computer and compared with the target value. When the sensed parameter deviates from this target value, the computer signals the actuator to "take action." This action is always in opposition to the sensed error (negative feedback), so the machine tends to correct or restore the parameter back to the target value. The machine constantly compensates, even when its target is moving, or after external conditions force it off the target.

This is not just a trivial, specialized aspect of machine design. The equations are the same and the results are the same whether it is a mechanical, electrical, or chemical process, and whether it occurs in machines or living creatures. Every living thing, from trees to human beings, employs this process for thousands of different applications. The closing of the iris of the eye in response to light, the regulation of hormones in the blood, and the ability to pick up a delicate object, these are all accomplished through the feedback control process.

What I have just described is well established and universally accepted. Still, it seems to be worlds away from explaining the complicated behavior of animals and certainly human beings. However, I have yet to discuss the full implications of feedback control in complicated robotic devices. I think once the "behavior" of these machines is understood the correspondence with the most complicated animal behavior will be obvious.

An arrow (or a bullet) traveling toward a target is subject to several factors that will cause it to miss the aiming point. Gravity pulls it downward, winds may move it laterally, or unsymmetrical construction may take it in any direction. Some of these can be compensated for, others cannot. Once an arrow leaves the bow or a bullet leaves the muzzle, the trajectory is uncompensated and the projectile strikes where it will.

A guided missile, or a "smart" bomb, uses sensors and the feedback control process to carry it directly to the target in spite of aiming errors, unknown winds, and even construction errors in the projectile itself. This ability to relentlessly seek and follow even moving targets seems pretty amazing, but it is nothing compared to that of more complicated machines. The level of "intelligence" and vast number of feedback control system operating simultaneously in an intercontinental ballistic missile (ICBM) makes its "behavior" much closer to that of a living organism.

ICBM’s have three distinct flight phases (ascending, free-flight and reentry) in which many different feedback control systems provide many varying functions to achieve the ultimate goal of delivering a warhead to the target. In the ascent phase, little or no consideration is given to where the target is located. The immediate problem is to compensate for loads imposed on the vehicle by winds or bending dynamics. When the vehicle gets through ascent phase, it may be far from its intended trajectory because of this compensation or unanticipated performance of its rocket engines. It may not even be "aware" of its ultimate mission until it reaches free-flight phase. Now it is necessary to change the trajectory back toward the target so that it will be in the box (within range) during reentry phase. It may have to avoid countermeasures, or maneuver to reentry attitude, but the primary goal is to be in the box at reentry. After reentry, the focus is on the target. Loads, over-heating or more countermeasures may still be considerations, but the overriding goal is to be on the target at impact.

Hundreds of feedback control systems operate simultaneously to control every aspect of the missile’s flight from liftoff to impact. Together they follow a complicated set of rules that vary from one flight phase and one set of environment conditions to another. This behavior is not preset or predicable before the flight occurs. It is not influenced in flight by any human hand. The machine makes the decisions alone by sensing what is happening, then taking action to compensate for whatever those conditions may be. This ability to react to whatever new conditions appear, with an ultimate goal in mind, is more than an analogy to behavior in living organism. It is exactly the same process.

Now that we have discussed the way sophisticated robotic machines behave, it is time to consider how even more sophisticated animals behave. The first thing to realize is that natural selection brings about the evolution of function as well as form. It changes not only beak, claw and eye, but the way these physical assets are used to assure survival and reproduction. Behavior evolves, and it evolves for the same reason and in the same manner as physical traits. A slight change in the preset goals in an animal’s brain may provide as much, or more, survival value than a change of physique. Over the entire span of life on earth, animals have acquired as many different behavioral traits as they have physical traits. As different as they have become, these functions have a common pattern, just as physical traits have a common pattern. The trick is to consider them one at a time, cutting through the fog of variation to see why they exist and how they work.

When we see a falcon striking a pigeon, it is a perfect analogy to a missile attacking an aircraft. The eyes are the sensors. The brain computes the error in the intercept path and sends a signal through the nerves. The nerves control the muscles, which control the wings, thereby changing the trajectory of the falcon. It is entirely a feedback control process. Of course a falcon is not programmed to attack every pigeon that it sees. It only has that goal imposed when its stomach is empty. The feeling of hunger is the signal that the brain receives from the stomach in yet another feedback process. Just as the ICBM goes through more than one phase to meet its ultimate goal, the falcon will go through a search phase, an attack phase and an eating phase to satisfy its hunger and reduce that error signal to zero.

The ultimate goal of every single living thing is survival of its DNA. If we fail to pass along our DNA, we have missed the target. Animals go through several life phases and it is not always obvious what the target is. The young are preoccupied with eating, growing and exercising. Young adults may seem to be more interested in competition, staking out territory, or showing off, but when the reproductive phase appears, mating becomes the overriding goal. In the family phase, care of offspring overrides the mating impulse. After all, survival of their offspring, the carriers of their DNA is the ultimate goal.

While animals may take years, or even decades, to meet their ultimate goal of reproduction and raising of offspring, there are usually more pressing problems. They have to stay alive. This means they have to find food, shelter and avoid predators. All of this is regulated and driven by the feedback control process. Some of these goals are much more urgent than others, and sometimes in the execution of one goal, another appears which overrides or conflicts with the first. For instance, the man who goes out to drain a swamp may decide his goal is of secondary importance when he finds himself surrounded by alligators.

It is this balancing of one goal against another, and the constant changing of goals based on condition of the individual as well as conditions in the outside world that makes animal behavior appear to be so complicated. When one goal is considered at a time, animal behavior is revealed for what it is—a feedback control process, pure and simple.

It may seem astounding that robotic motions based on preset goals could explain the intricate behavior of animals. However, when you consider how and why animals behave as they do, and how they acquired these traits in the first place, it is perfectly obvious that this is exactly what is happening.

Instincts are the long-range drivers of behavior in the animal world. Migrations, defending of territory, finding a mate, building of nests and care of young are all examples of instinctual behavior. Nature could not depend on animals to decide on their own that they needed to do these things. These traits had to be preset into every individual, so that they would feel they wanted to, or had no choice but to follow these rules. They are too important to the ultimate goal of preserving DNA to leave it up to the individual. Animals are driven to do these things to close the feedback loop and set the error signal to zero. When they do these things they are on target for nature’s imperative. This is just as true for mankind is it is for any other animal. We may feel that we do them for pleasure, for fulfillment, or to avoid the pain of failure. It is really the way that our feedback loop signals us to get with the program and fulfill our obligation.

Emotions are short-term drivers of behavior. They act over minutes or seconds rather than the years or decades required for instincts. Anger drives us to defend our territory, our food supply, or even our intellectual positions. Fear drives us into retreat when we sense that we are unequal to the threat. When a chipmunk comes out of his burrow and discovers that a stranger is helping himself to food in his territory, you can almost see the flush of indignation and anger. The brain immediately sends the signal to charge down and bite the adversary in the butt. That is the overriding goal until the intruder is driven away. Of course, if the intruder is twice as big, then usually before teeth meet the target, a signal is sent to change the goal. When fear overrides anger, the goal is to turn tail and make for the burrow.

Instincts and emotions are just as much functions of a feedback control system as any of those other that work every minute to keep us living and breathing. Every instinct and emotion can be identified with some particular controlling function in animal behavior. Humans still have all the instincts and emotions of their animal ancestors and those controllers are as strong and active as ever. We probably have more logical minds in some respects but that is a specialized ability like the power of speech or hands to use tools. These can affect the way we reach our goals, but the goals themselves and the drive to reach these goals are part of the feedback control function. The logical mind is a powerful tool for some applications, but the amount that it influences human behavior varies from none to almost none. Some people live from birth to death with behavior totally unaffected by the logic circuits they are carrying around up there in the frontal lobes. This is not a strange condition. It is the way that nature intends for things to work. After all, nature has been pulling the strings for us since life began. Who can we trust to do a better job?

Pain and pleasure are just little gimmicks that nature has devised to make sure that we follow its rules. Mother (nature) knows best and is not hesitant to give us the carrot or the stick since it is in our best interests. Pain is the feedback signal that instructs us to take our hand off a hot stove. It may also be used as a warning not to eat too many green apples, or to not lie out in the sun for too long. We can also have emotional pain or pleasure when the occasion calls for it. Nature is not too concerned about our comfort or our emotional state. As carriers of the precious DNA, it often gives us a hard lesson in how to take care of ourselves so that we can get on with the important business of survival and reproduction. When we are eating good food, enjoying a comfortable environment, defeating an adversary, mating, or caring for children, it gives us a little reward of pleasure. We are doing what it wants done. The feedback loop is closed and we are on target.

The easy way to understand human nature is to look at it from a biological perspective using feedback control theory. People are born with slightly different set points in their life goals, just as people have slightly different set points for normal temperature or blood pressure. People also have slightly different gains and damping coefficients that make them more or less driven toward their goals and more or less stable in their behavior. This sum total of all set points and gains is what defines individual personality. This is not an error in design any more than variations in physical size or shape are errors in design. Nature may abhor a vacuum, but it loves variety, whether it is in form or function.

People will doubtlessly go on believing that human nature is influenced by supernatural forces of good or evil, by good or bad parenting, or by too much riches or poverty when they were growing up. The problem with these theories is they lead you down blind alleys, into self contradictions, and cannot explain the facts of human nature as we see them around us everyday. Knowing that people learn from experience and pick up habits from their associates leads to the illusion that personality traits, character, goals and ambition are re-programmable. It is hard for people to admit that their loved ones are functional failures and there is nothing they can do about it. It is even more difficult for those whose professions are devoted to such re-programming to admit that their efforts are worthless and hopeless. This is not a desirable state of affairs, but many of the things we find in nature are not as we would wish them to be. The only way to get by is to accept them for what they are and find a way to work around them.

The more we discover about the fine details of the structure of matter, or the fundamental laws that govern the universe, the stranger they sound. Perhaps the natural world is too complicated for the human mind to ever comprehend. Perhaps speculating about human nature will not do us any good in the long run. Still, it is something that a few of us are compelled to do. Curiosity is a facet of our personality. We seek pleasure in knowing. We must drive the error signal of doubt to zero.