gen_47.1.gif WelcomeAboutThe Booke-mail me

Chapter 7 - Free Will

Moral choices require volition which is an emergent property of thought, but the determination of whether an entity has free will is a relative characterization, not an absolute one.

Every action has a measure of goodness to it if the welfare of some entity can be affected by it. But morality is not meaningful to a rock falling downhill. If falling downhill causes the rock to shatter into small pieces, then the action of rolling downhill is not in the well-being of the rock qua rock, since it now has ceased to exist. In terms of the constituent parts of the rock, the action is neutral.

Morality comes into being when there is conscious choice involved. A rock falling downhill is not making a moral choice; it is acting in its own nature.

Living creatures - plants and animals - can make choices. That is, there is some internal representation of the entity's well-being that is referred to in the determination of the action to take in response to the current environment.

A choice was initially defined an action that can be good or bad, depending on the well-being of any entity affected by that choice. This definition needs to be refined, since the action of rolling downhill in a way that causes a rock to shatter is a choice by this definition. To call this a choice by this rock stretches the definition of choice beyond what is reasonable.

The definition will be extended by requiring that the entity performing the action must have access to some representation of the well-being of some entity (not necessarily itself) affected by the action. This representation is used in decisions that serve as a precondition for the action. This means that there is some measurable condition that derives to some characterization of well-being for an entity, such that a change in this condition would result in a different action. The entity is said to react to this condition. In a simple creature such as a sunflower, the action of turning towards the sun is a result of the determination of where the best sunlight is, a condition that is a measure of the plant's well-being. In that sense, the sunflower chooses to turn towards the sun, or it reacts to the sun moving across the sky.

Note that the word choice is quite often used to mean a decision made by conscious deliberation. This is not the sense that is being used here. If a conscious choice is being considered, terms such as judgment or volition will be used to describe it. Choice by itself refers to actions that are simpler or more basic - all they have to do is affect well-being with a determination of that well-being part of the preconditions.

For simple plants and animals, these conditions exist as measurable quantities and configurations of the animal, but the representation of well-being and the resultant choices that are made are qualitatively different from the decisions of more complex animals. Many animals have a nervous system, part of which is reserved for the representation of the well-being of that organism. Pleasure and pain are terms used to identify some of these neural representations.

Animals choose to react to pleasure and pain, but only the most advanced creatures use judgment. Judgment requires that the organism not only identify the pleasure or pain that the organism is experiencing, it also requires the organism identify to some extent the alternative actions that are available to it. The sunflower does not use any judgment to choose to turn towards the sun - the action is automatic. Similarly it is unlikely that a fish uses judgment regarding whether to school or not.

When we get to the mammals, the situation is not so clear-cut. It is obvious that almost all normal humans use judgment. But the question of whether a dog makes a judgment when choosing the time to have its dinner is a question that would take a psychologist to answer. A conditioned response such as salivation takes no judgment on the part of the animal. But there are other choices that an animal can make that could arguably be said to include judgment - that is the animal's decision involves not only a determination of well-being, but also a determination of the expected change in well-being an action would result in.

An example of this is learned helplessness. A dog placed in a cage with a barrier in the middle where either of the two floors can be electrified, will soon learn to jump to the other side to avoid a shock. By itself, this could be considered to be simply a conditioned response that requires no judgment. But if both sides are electrified, then this action becomes useless and the animal stops jumping. This results in the animal's exhibiting signs of distress. Again, if the animal just stopped jumping, this could be argued that the animal is capable of choice without judgment - it chooses to stop jumping because this wastes energy without helping. But the resultant distress the animal experiences makes a compelling case that the animal has some notion of its own condition, a necessary component of the process by which it forms judgments.

This is the basis of intensional action. Intensionality requires both reflection and adjustment. The adjustment for the dog occurred when the situation changed and the animal found that the avoidance was not possible anymore. Intensionality requires a certain amount of self-consciousness - the ability to reflect on the state that oneself is in and to make these adjustments based on that reflection.

As far as we can tell, though, human have a qualitatively greater level of consciousness than other creatures, and therefore a greater degree of volition. Where does this come from? Is there some special organ, or part of the brain that gives rise to this heightened consciousness or is it some emergent phenomena of thought?

Although there are most likely significant differences in human brain structure leading to the uniquely human level of consciousness, and thus a physiological explanation of consciousness may be achievable, I would argue that the experience of consciousness that most humans share is an emergent property of the mammalian brain.

First, I would make the case that consciousness comes from the ability to learn. The proof of this is not difficult. It is only necessary to point out that all people do not start out being born conscious. We tend to have conscious self-awareness by some time around age three or so. At this point there either must have developed some complex of neurons that caused self-awareness to arise, leading to a qualitatively different mode of functioning in the brain, or that we have accumulated a certain body of thoughts by learning them. Even if there is such a neural complex, it would appear at the same time that language acquisition, abstract thoughts and ideas and other traits arise in the child. These behaviors and ideas are learned. Actually, self-awareness is useless if one can't learn, because there would be no change in behavior. Therefore learning is necessary for the development of consciousness.

Without speculating about physical changes, the way consciousness arises can be thought of by using the metaphor of a nuclear reactor. A pile of uranium contains atoms of fissionable material. These atoms, after having absorbed a suitably condition neutron go into an unstable state an split in two, generating more neutrons. There is a qualitatively different state between criticality and subcriticality. The subcritical pile is inert, benign. A critical pile is hotter, more radioactive, and contains larger amounts of non-uranium atoms. It has characteristics that were too diffuse to be a part of the subcritical mass.

This is probably what consciousness is all about. Any collection of neurons capable of registering pleasure and pain has a subcritical consciousness, in a way that inert materials are not capable of. But the mammalian brain has a sufficiently high degree of thought processes that give rise to thoughts, which in term learn other thoughts, just as one fission gives rise to two or more others in close enough proximity to suitably conditioned neutrons.

Animals have a wide range of consciousness, but probably even the most advanced mammals outside of the primates are subcritical in their thought processes. Therefore their level of consciousness is qualitatively different from people.

This way of looking at consciousness may help explains how consciousness fades away when going to sleep and disappears in deep sleep or moments when one is zoned out. People who are sleepwalking can commit good or bad actions but are not morally responsible because they had no powers of volition. Even if their unconscious states come about because certain parts of their brain are shut down, this may mean that they are only capable of a series of thoughts that remain subcritical using this nuclear reaction metaphor.

Although an interesting metaphor, defining the point at where consciousness begins is difficult. The attempt to pin down this point may not be useful, though. Most creatures are well above or well below the line. Even a sleeping person, who goes from a waking state into a deep sleep, then into a dreaming state, then back into a deep sleep spends most of their time either conscious or unconscious and very little time in the transition between the two. A two to three year old human is typically the only creature who lives at the line for long. It is possible that a chimp can be at this level, also. It has also been speculated that dolphins can achieve this also, but it would be hard to give a definitive answer.

With these views of choice and consciousness, we can now look at free will. Free will has traditionally been considered to be an all-or-nothing proposition: either an entity (usually human) has free will or it does not. There are a number of types of free will that have been considered, typically characterized by whether the choices of the entity in question are nondeterministic or not. For instance, some consider humans to have free will because the choices they make have an essentially nondeterministic, unpredictable component. There is an essentially random character to the choices. Others consider an idea of free will that is compatible with an essentially deterministic view of the choices that entities make. What both views have in common is that the entity does not employ a simple mechanistic thought process. Even for the believer in compatibilism - a free will based on a deterministic thought process - that thought process is usually non-trivial, involving considerations of the immediate environment, internal mental states and future expectations.

I shall instead argue that free will is a relative judgment call. As such it depends on both the viewer and the entity being viewed. It is not an absolute determination. The only reason free will has appeared to be absolute was because the only entities who have considered the question of free will - that we are aware of - are our fellow humans. This makes it appear absolute, because there is no relative frame of reference except the variation among people.

Although the field of Artificial Intelligence has repeatedly met with failure, there is, to our knowledge, no theoretical reason that explains why there can never be an electromechanical computer that could attain or surpass human intelligence. Some progress has been made, but so far there has not been a theoretical design that has led to a computer capable of thinking on the same level as a human over any but a very narrow range of abilities. Despite this failure, the rest of this discussion will, for the sake of argument and some examples, presume that such a machine is possible.

I first point out that, with this relative viewpoint of free will, there is no free will if there is no observer, which implies a certain level of consciousness. The consciousness is required to provide a reference point by which to judge the range of choices made by the entity under observation to the range of choices made by the observer. Therefore, the question of free will does not come up for plants and simply animals. They are incapable of evaluating free will - it does not enter their mind, such as it is.

We humans have free will relative to the viewpoint of other humans. Expressed in the alternative manner, every organism has free will as far as it can determine, if it is capable of making that determination. Given that you yourself have free will, you look at most other humans and determine that they have free will also. There are exceptional cases, but they are determined by acts of introspection and empathy. That is, a person in a vegetative state may be considered to have no free will because observation of that person compared to introspection of oneself leads one to conclude that the ability to make significant choices is profoundly impaired.

If an entity is completely deterministic to an observer, then that entity has no free will. This is a basic precondition the absence of which leads to a judgment that denies free will. It is necessary to show that the most humans cannot have their basic actions be reliably predicted by a human observer, even if their actions are completely deterministic, and thus to other humans, the typical human has free will. The argument is based on a computational complexity argument.

When a person views an object, either living or a machine, this observer abstracts a certain level of behavior from the object's actions. That is, if a human watches a dog play a game of catch, the person does not describe the actions of the dog in all its minutia such as the precise leg movements the dog makes. Instead, the observer abstracts the condition that the dog is chasing the ball.

These abstractions characterize the object as a member of a set of similar objects. I can say that a human has certain characteristics that make a human different from other animals. But we claim that humans have free will because there is no set of abstractions that serve to characterize the ability of humans to make choices except in the most superficial of ways.

This failure to abstract the essential qualities of choice is not necessarily due to a stochastic component of the choices that people make. It can come entirely out of deterministic processes, although they are incredibly complex. Recent work in chaos theory shows that many non-linear systems exhibit chaotic behavior very dependent on initial conditions. The gross qualities of this behavior can sometimes be characterized in general terms, but the detailed moment-by-moment behavior may appear unpredictable. And in some chaotic regions even the gross properties are beyond analysis. Some of these systems are termed "trap-door" functions: they are very easy to compute in one direction, but incredibly difficult to compute their inverse.

Human choices are type of deterministic behavior. The choices that a typical person may make can show an overall regularly from day to day, although that can disappear when that person goes through life-changing events. The choice function can be considered to be a trap-door function. The individual choices may be the result of an easily-explained rationale, but the attempt to arrive at that rationale from the simple listing of the choices made can be impossible to compute efficiently.

Any attempt to make such an analysis is bound to fail due to the capability of humans to anticipate the reactions of other humans to their activities and adjust their actions in reaction to this anticipation, if a conflict in goals or desired outcome exists between them. As we will discuss in a later chapter, this ability of humans leads to an essential inability to decide what a person will do in a certain circumstance.

To abstract the qualities required to make predictions of what a person will do is often beyond human analysis. It may be reasonably predicted that this person will have a meal sometime in the next day or so, but unless the person's diet is forced to be limited, the particular menu that day may be off the mark. The appearance of free will arises when the most efficient way to make such a prediction is through the use of simulation - that is, create a copy of the person and sit down and watch what the copy will do. Unfortunately, it is not currently possible to create such a detailed, usable copy of a person with our current state of knowledge. So at present, we can only observe the original and see what that person will do. If the most efficient way to predict a entity's behavior and the choices they make, then that entity has free will.

This view of free will implicitly requires an evaluation that is relative to the entity making the determination. For a human or a computer as powerful as a human, a lizard may have no free will, but another human does. Whether or not a dog has free will is debatable. A dog is able, to a limited extent, anticipate what a human may do in reaction to what the dog does, but this level of anticipation is limited both in time and degree of sophistication, limits that humans do not possess. For a theoretical computer more powerful than the mental processes of most mammals, a dog may have no free will but a human has. For a computer with godlike powers, even a human has no free will.

This relative view of free will clarifies the relationship between God and man in terms of free will. To a human, another human has free will, but to an omniscient god a human does not. This is true even if humanity's actions are not willed by an all-powerful deity. Since free will is a matter of evaluation, omniscience alone is enough to lead to a lack of free will in God's eyes.

The typically absolute interpretation of free will naturally leads to a misapplication of the consequences of divine predestination found in some Protestant denominations. The presumption that an omniscient god can perfectly predict the actions of any given human says nothing about the way in which a human should function. This is due to the fact that predestination and a lack of free will is obvious for an omniscient god, but the reduced capacity of human reasoning leads to a presumption of free will and the necessity of making appropriate moral choices.

This relative interpretation of the concept of free will is compatibilist - that is, even if there is a recognition that the world is essentially deterministic, free will is compatible with this determinism. Even though there may be a non-deterministic component to the universe, this nondeterminism does not play a significant role in the typical choices that humans make.

The opposite of compatibilism can be addressed with this relative model. The incompatible "requires that there is a plurality of futures open to me consistent with the past (and laws of nature) being just as they were." [Stanford encyclopedia of philosophy] The incompatibilist viewpoint requires two assumptions: first that there exists a basic indeterminacy in reality, and second, that this indeterminancy plays a nontrivial part in the decision making process of entities such as humans. Otherwise, incompatibilism is an illusion brought about by the impossibility of identifying all of the deterministic processes. The relative view of free will implies an identification of an entity viewing another as if that viewed entity has a plurality of futures even though no such thing actually exists. This plurality is not inherent in the nature of reality; instead it comes from the inability of the observer to reliably rule out significant alternative outcomes. As the depth of analysis of the entity's future possibilities increases, an essential determinism leads to the eventual elimination of alternate futures as possible until just one is left. That is why the determination of free will disappears as the complexity of analysis increases.

This complexity also increases within humans. As we learn more and more about psychology and sociology, we identify less and less situations in which an individual is practicing free will. Although the potential exists that a state can be reached in which no human could be judged to have free will, in practice this is likely to be unobtainable.

One argument against a purely deterministic view of nature is from the discoveries in the twentieth century that the laws of physics in the smallest scale exhibits and essential indeterminancy that can only be expressed using probability functions. The best known expression of this is the Heisenberg Uncertainty principle of Quantum mechanics. Quantum mechanics could supply an indeterminancy that is potentially an important influence on human decision making.

But it does not appear that way when looked at carefully. Quantum mechanical processes, even in a world controlled by chaos theory - which is profoundly affected by minuscule changes in initial conditions - affects the range of situations available for a human to make moral choices, but does not appear to be an important factor in the way a human makes those choices. It appears that all of the probabilistic quantum fluctuations may operate on such a low level that they do not enter into the day-to-day decision making processes of the typical human. Rational mental processes seem to follow deterministic logical rules, that are either internally consistent, or if they are not have a rationalization that can be described by a determinable syllogism. Even emotional mental processes have a distinct physiological process that can be described in terms of neurons firing in certain regions of the brain, where the quantum fluctuations more often than not cancel out. Therefore human decision making tends to be composed of reasons and desires that have an almost discrete objective existence that is as little conditioned on probabilistic events as the chance at a word on the printed page could, through quantum mechanical processes alone, rearrange themselves into a piece of gibberish. For example, the operation of today's computers is potentially affected by quantum mechanical events, but the computers are designed so that this is minimized. It also seems likely that these effects do not play a part in human reasoning. Therefore, although quantum mechanical processes can make minute changes to the possible choices that a human can make, in general, it is unlikely that these processes significantly affect outcomes.

Besides noting that there is a significant difference in the judgment of the amount of free will an entity has based on whether the observer and the observed are qualitatively different entities, such as a lizard, a dog, a human or God, it is always possible to draw finer distinctions even when the comparison is between entities of the same type. These finer distinction can come from the relative intellectual capacity of the observed entity versus the observer. An extreme case is between that of an adult and a child, but a distinction can also be drawn between the capacity for free will between an educated person and an uneducated one, or expressed in terms of a person's life experience instead of their years of schooling, between a wise person and an unwise one.

The other difference in judgment is of course in the determination of the situation that the observed entity is found relative to the situation of the observer. It is obvious that the determination of the amount of free will an entity possesses is different if the situation that the entity is in is understood with a deeper level of knowledge of its implications by the observer than the observed. This is always a factor in culpability. This determination of free will can change depending on the situation of the observer or the situation of the observed. There is an essential symmetry changing the determination of free will such that a change in the observed has a corresponding possible shift in the situation of the observer.

The more educated or wiser, a person becomes, the finer the distinctions can be made. A deeper appreciation of the complexities of the situation leads to a change in the determination of whether the actor acted with free will or not and is therefore culpable. Most often this increased knowledge of the situation leads to a determination that the actor had less free will than a more superficial analysis shows, but this is not always the case. It can also happen that a deeper analysis can show that the entity actually did have alternatives available that the entity was not aware of. This means that the observer would declare that the observed acted with free will even though the observed would claim that the action was the only alternative open at the time.

The determination of an entity's free will is not just an abstract thing. The notion of free will is important because it affects the way that individuals react to the actions that an entity takes. The reactions that other people have to an entity's actions sometimes take the form of rewards and punishments, but there are other reactions that involve the notion of free will besides this. For example, the expectation that a person can learn from their actions is dependent on the degree of free will the entity is thought to possess. Therefore, the amount of teaching given to this entity and the depth of this teaching, such as expressing this knowledge in terms of motivations depends on the amount of free will the entity is judged to have3, especially since free will is partly determined by the entity's capacity to reason with self-awareness. This is true both for humans and for teachable animals, such as most mammals.

But the question of rewards and punishments is usually the starting point for determining how to react to another's actions in light of their capacity for free will. An unfortunate outcome can be excused or even met with sympathy if the action is deemed to be beyond that person's capacity to alter. On the other hand, the determination that a person acted with willful disregard to what is right will not only cause a harsher punishment to be meted out, but the overall judgment of that person's character will suffer as a result.

The determination of free will is considered in light of the number of options available to a person and the ability of that person to choose amongst those options. With an absolute notion of free will, this determination is complicated enough. With a relative notion of free will, not only does the mental capacity of the actor come into consideration, but also the mental capacity of the observer making the judgment. This two body problem leads to a complicated dance of reasoning that can lead to different results, especially as the mental capacity of most living creatures is constantly in a state of flux.

But because free will is relative, systems of rewards and punishments are relative also. Punishment has two main purposes - deterrence and retribution both of these are dependent on a person's degree of complicity and this depends on their volition. Even if complicity can be objectively determined, volition never is.

One way to reduce the complexity of making these judgments is by reference to a standard. Standards still exist even when the determinations are relative. It was already pointed out that in the determination of what is good, there is an absolute standard even though what is good is relative. This absolute standard is expressed in the Golden Rule and is derived from an application of empathy. The same process applies to judgment. The reaction to a person's actions, the reward and punishment, requires empathy on the part of the person doing the judging to determine what capacity for free will the actor had in that case. But this is not entirely enough because every person's capacity for free will is dependent upon their situation. So the capacity for reasoning empathetically must take into consideration the typical situation the actor is in, not the situation of the person doing the judging. One way to objectify this is to postulate some hypothetically typical person, and compare the actions of that actor to this hypothetical.

The starting point of this evaluation is for the person making the judgment to consider themselves a typical person, and then make the judgment based on a sense of empathy. This requires that the observer take into account the difference in the situation that the observer and the observer find themselves, especially since such judgments are usually made based on events that have happened in the past. This means that the situations differ greatly, since the observer knows the actual outcome when the actor probably was not able to predict this with any reliability. It takes a fair degree of empathy for the observer to ignore these differences and to put themselves in the same place as the actor with the same lack of knowledge.

This comparison to the typical person forms the bedrock of society's laws. But comparison to the individual is the basis of most people's judgment of others. This is not a good basis of defining and applying the law, though.. This is because our comparisons are usually distorted by the high self-image that the typical person holds of themself. This means that empathy must not be given in comparison to the observer of a moral act, but made on the basis of contemplating a third, theoretical construct - the image of the typical person.

When, in the determination of the responsibility that a person has for their actions - whether that person be punished for an evil act or rewarded for being virtuous, the point must come when free will is presumed, if the reward or punishment is a response sufficient in itself - that is, the response is made for no other purpose than to reward or punish. This determination can be made by comparing this person to the typical person. If a typical person would have acted more benignly in the same situation, then the person being judged would be considered to have done something wrong. If the situation was such that the typical person would have done no better, regardless of the bad consequences of the act, then the person is judged not to be responsible. We also declare someone not responsible if the outcome was predetermined - this happens for people such as children or with impaired mental capacity.

If the application of reward and punishment is done for the purpose of correcting future behavior, then it appears that free will is not as important a consideration in whether or not to mete out that response. In this case, the determination of free will is not as important as the determination of whether the future behavior of the actor will be changed in a way that is desired by the observer due to the reward or punishment. If the recipient of the reward or punishment is deemed to possess free will, this means that the observer is less likely to successfully predict that the application of this corrective action will lead to the desired result.

One unfortunate effect of this indeterminacy in the face of free will is that the person who metes out the judgment may tend to over-emphasize the degree of the reward or, especially the punishment in the light of free will, because the result is less predictable. That is, knowing that the actor had limited free will and knowing the effect on the future behavior of the punishment, the punishment will be more finely calibrated due to the degree of knowledge. In ignorance, an attempt will be made to further guarantee success on correcting the behavior of the actor by increasing the degree of punishment.

That question arises then, to what degree does one punish a transgressor who grew up in a deprived household? There is often a tendency to overestimate the degree of culpability of such a person, or an overestimation of the effect of punishment or reward on future behavior. This is due to the tendency to neglect the difference in situation that the observer came from relative to that of the transgressor. To make an appropriate response, this decision is made in terms of the average person - whether they would grow up in the same situation would make the same transgressions also. This is not to excuse the bad behavior of someone who grew up in a bad situation from the consequences of their actions when they ought to know better. But it does argue that a person of privilege, who is given the opportunity to truly know the difference between right and wrong, should by rights be judged more harshly than the person who does not have the opportunity to see those differences.

One other consideration that must be made is the recognition that there are different levels of entities, and that each of these levels can be judged on their own merits. That is, the situation that an individual finds themselves is no more than the characteristics of that part of the society in which they were brought up. It is reasonable to grant a certain amount of free will to a society as a whole, since that society manifests a certain level of volition that results in an emergent behavior that is different that the individual behaviors of the people that make up the society. That is, the volitional actions of the individuals can lead to a qualitatively different volition for the society that aggregates these actions. This means that it is possible to judge and reward or punish a society independent of the judgment of the individual in the society.

This observation leads to the conclusion that praise or blame for an individual's actions can be apportioned between the individual and the society of which they are a part. An example of this is the matter of racial prejudice. If a society as a whole is prejudiced against a particular group of people, it is reasonable to condemn the society more than the individual, although both must share the blame to some extent.

It is interesting to note that if there is a god that is all loving and omniscient, and the rewards and punishments are done for no other purpose than reward or punishment itself, then that god can certainly create a heaven or hell to separate the sheep from the goats. On the other hand, if the purpose of reward and punishment is to correct, then the all-loving nature of that god implies that god must be a universalist.


We have predicated the ability to make moral choices on the existence of consciousness. It is this consciousness that makes the moral choice volitional. The question arises though, whether these choices are predetermined or due to free will. We have seen that free will is not an absolute measure - it is instead a relative judgment based on the different mental capacities of the observer making the determination of whether free will exists or not relative to the mental capacities of the observed entity whose will is being judged. This determination must also take into account the situation in which the entity finds themselves. The degree of free will determines the culpability of the entity making a moral choice - an entity that can is judged not to be able to make a free choice cannot be held to blame by the entity making the judgment. Finally, we note that the blame can be shared between individual entities and the larger entity that defines the situation the individual finds themselves.