Responding to Moral Apologetics

In comments on the essay The Case Against Moral Realism, I challenged someone (Stephen Lindsay) to define good and evil, or in other words, to define moral value. In response, he wrote a defense of belief in objective morality: Three Perspectives on Objective Morality.

In the essay, he gives a definition of moral rightness and wrongness:

• Right (in the moral sense) - behaviors, ideas, philosophies, etc., that are thought to overall positively impact individuals and society. Wrong is the opposite. There can exist behaviors, ideas, and philosophies that are neither right nor wrong.

However, this definition does not actually define moral value. It does not explain what moral goodness and badness (or rightness and wrongness) are. It presupposes moral value.

To call something “good” or “bad”, in a moral sense, is not a description. Moral value is normative. To call something “good” (in a moral sense) means that it ought to be, and that we should try to create and maintain it. Likewise, to call something “bad” (in a moral sense) means that it ought not to be, and that we should try to prevent it.

Moral value is supposedly normative from a cosmic perspective, and it is also normative to us, as subjects. It is both objectively and subjectively normative (which makes it a very strange concept).

You cannot define moral value simply by giving a description of what you believe to be morally good and bad. You have to explain why those things are normative. Also, the objectivity of what you believe to be good or bad does not establish the objectivity of moral goodness and badness. For example, murder is an objective event. It does not follow that the moral badness of murder (if you believe it to be morally bad) is objective.

To defend moral realism, you would need to satisfy the following requirements:

1. Ontological objectivity: Describe the objective properties of an event/outcome that make it morally good or bad.
2. Objective normativity: Explain why the description in (1) implies objective normativity.
3. Subjective normativity: Explain why the description in (1) implies subjective normativity.
4. Fit to intuition: Show that moral goodness and badness (as defined) fit the moral intuitions of most people. If there are systematic differences, explain why they exist.

Both (2) and (3) require bridging the is | ought gap. The moral realist must derive normativity from a description of objective properties.

In most cases, the moral realist will leap over the is | ought gap by begging the question implicitly. He will tacitly assume some moral principle, such as “We should be altruistic”, and then appeal to that unstated assumption.

We can see that leap in the definition given by Stephen Lindsay. He simply assumes that we ought to do what is positive for individuals and society. No explanation is given for why this constitutes moral value. He does not explain how it is objectively or subjectively normative.

Stephen doesn’t define “positive”, so it’s not clear what he believes to be good for individuals or society. He presupposes those types of value.

There are different types of normativity. For this essay, the important distinction is between “good for” and “good”. To avoid confusion, I will call the first one “p-value” and the second one “m-value”. P-value is what is good for an entity, which could be an individual person, an organism, a group of people, etc. M-value is what is morally good. The concept of moral value presupposes p-value, as do many other concepts, such as “self-interest”, “selfishness”, “altruism”, “benefit”, “harm”, “help”, “punishment”, “reward”, etc.

I don’t want to take p-value for granted implicitly, and I don’t want it to be confused with m-value. However, I also don’t want to spend too much time talking about p-value and the various ways that it could be defined. That’s a topic for another essay. But we need to make the distinction between p-value and m-value, otherwise we will become confused, and probably go in circles.

M-value is presumed to be objective in two ways. One is that it does not depend on a valuer. Moral goodness and badness are presumed to be objective properties of events or outcomes, not subjective attitudes or judgments. Of course, individual subjects can make moral judgments, and those judgments are subjective, but they are supposedly about something that is objective. For the simplicity of this essay, let’s assume that p-value is also objective in this way. M-value is also objective in that it is not relative to an individual or entity. Moral good is not “good for X”. It is just “good”, period. Likewise, moral bad is not “bad for X”. It is just “bad”, period.

For example, suppose that Tom killing Joe has negative p-value for Joe and positive p-value for Tom. In other words, Tom benefits in some way, while Joe is harmed. The event has zero p-value for Mike, because he receives no benefit or harm from Joe’s untimely death. So, the same event has different p-values for Joe, Tom and Mike. By contrast, if murder is morally bad, then Tom killing Joe has negative m-value, period. There is only one m-value for the event.

Although there is only one m-value, which does not depend on the valuer, the m-value is also (supposedly) normative to every subject. Mike should try to prevent Joe’s murder, even though Mike would not be harmed by Joe’s death. Tom should not kill Joe, even though Tom would benefit from Joe’s death.

In light of that distinction, let’s return to Stephen’s definition of moral right and wrong. He is saying that moral good is whatever benefits humanity as a whole, and moral bad is whatever harms humanity as a whole. In other words, moral value is the net p-value for humanity. At least, that seems to be what Stephen is trying to say, although he is not very clear.

We can define the p-value of a collective as the sum of p-value over the members of the collective. That’s not the only way that we could define it, but it’s the simplest way, and it seems to be Stephen’s implicit definition.

Stephen does not define p-value for individuals. He takes it for granted. For the sake of argument, let’s assume that we accept his implicit notion of p-value, whatever it is.

Would the p-value of the human collective, defined as the sum of p-value for all human beings, constitute m-value? Would it satisfy the requirements for a definition of moral value?

No. A summation does not bridge the is | ought gap.

Let’s assume, for the sake of argument, that his implicit definition of p-value satisfies the first requirement: it is an objective property. That’s dubious, but let’s assume it. Would requirements 2 and 3 be satisfied?

No. Stephen hasn’t explained why the p-value of humanity is objectively normative, or in other words, why it is m-value. He has not explained why the p-value of humanity is subjectively normative to me or any other subject. Why should I act toward the p-value of humanity, instead of my own p-value, or something else?

Suppose that my p-value is subjectively normative to me: I subjectively view my p-value as defining what is good for me. Would it follow that the p-value of the human collective (or any other collective) is normative to me? No. In fact, the opposite is true. If my p-value is subjectively normative to me, then the p-value of a collective is not subjectively normative to me, because it would probably conflict with my p-value.

The essential properties of moral value cannot be derived by summing p-value over a collective. Stephen has not given a satisfactory definition of moral value. He has simply expressed his unexamined assumption that acting for the greater good (collective p-value) constitutes moral value.

He has not even argued that it fits the moral intuitions of most people. In fact, it doesn’t. There are well-known thought experiments, such as the utility monster and Omelas, that show the counter-intuitive implications of moral utilitarianism (summing p-value to get m-value). Moral intuitions are not just about the greater good, although they might be in some cases. They often involve some notion of fairness, which is not captured by a simple sum.

Stephen seems to have taken the collective for granted as well. Why humanity? Why not all sentient beings? Or why not something more exclusive? Why not my race, my country, my friends and family, or … just me?

✦ ✦ ✦

Now, let’s consider some ways that a moral realist might try to derive normativity from objective properties. These correspond to sections in Stephen’s essay, but I will describe and refute them generically.

• Moral value is God’s will, and God (as a cosmic subject) is objective.
• Moral value is implicit in human nature. We are naturally altruistic, because of empathy and compassion, or some other “moral sense”.
• Moral value is implicit in the cosmic order. Moral goodness is the cosmic telos, or part of that telos. Thus, it is normative to us, because we are part of the cosmos.

Let’s consider each one.

In the first one, God is an objective (or at least cosmic) source of moral authority. His will is a cosmic standard by which things can be judged. Of course, there are many problems with this.

First of all, it assumes the existence of God. If the moral realist is claiming that moral value is objective because it is God’s will, he would have the burden of proving the existence of God.

As an aside, theists often claim to have answered one question by replacing it with two questions. For example, the theist might claim that positing God explains why the universe exists. But of course, it doesn’t. Even if we accept the premise that God exists, we now have two problems of explanation instead of one. We can ask why God exists, and why God created the universe. Nothing is explained by positing an inexplicable, omniscient being.

If we assume that this option satisfies the first requirement (ontological objectivity), then you could argue that it satisfies the second requirement (objective normativity), simply by associating the cosmos/objectivity with God’s subjectivity. This is also dubious, but arguable. However, the third requirement (subjective normativity) would remain unsatisfied.

If X is God’s will, it does not follow that X is normative to me. “X is God’s will” is a descriptive statement. Even if I believed in the existence of God, I could still reject his will as normative to me. Why should I care what God wants? Why should I obey his will? It is not my will.

The notion of God does not bridge the is | ought gap.

There is also the problem of knowing God’s will. Most religions place God off-limits to human observation and understanding. God is unknowable, unfathomable, etc. That’s convenient when you are trying to dodge questions, but it’s not so convenient when you want to make a positive claim. If moral value is God’s will, then moral value is unknowable. How would we recognize good and evil using this definition?

It’s worth pointing out that, according to the Bible, God has done and said many things that are now considered evil. God destroyed humanity 1.0 in a massive flood, sparing only Noah and his family (and, it seems, forcing them to commit incest). God told the ancient Hebrews to commit genocide.

Completely destroy them—the Hittites, Amorites, Canaanites, Perizzites, Hivites and Jebusites—as the Lord your God has commanded you.

[Deuteronomy 20:17, Bible, New International Version]

This does not fit modern moral intuitions.

More generally, there is the problem of evil, which is a major challenge to religious belief. Many passages in the Bible point to the amoral nature of objective reality — and thus, presumably, of God.

He causes his sun to rise on the evil and the good, and sends rain on the righteous and the unrighteous.

[Matthew 5:45, Bible, New International Version]

Now, let’s consider the claim that moral value is implicit in human nature. That’s a typical humanist belief: that humans are innately altruistic, due to empathy and compassion. Supposedly, there is a moral “sense” in the brain, which is the foundation of morality.

Having an innate moral standard would not make moral value objective. Moral judgments would reflect this inner standard, not an external, cosmic standard. We would not be perceiving objective good and evil, in the way that we perceive a tree or a rock. Good and evil would not exist independently of our minds, as the tree and the rock do. We would be projecting good and evil onto things. Moral value would be no more objective than personal value.

If I want an ice cream cone, that value is not intrinsic to the ice cream cone. The value comes from my brain. It is a value judgment, not a perception of reality. Likewise, if you judge the Rwandan genocide to be evil, you are projecting negative moral value onto an objective event. The event really happened, but the moral value is not an objective property of the event. It was assigned to the event by your brain. It exists within your mind, not in objective reality.

If humans had an innate moral standard, then it would be subjectively normative in a way. If you inherited normal moral instincts, you would have normal moral feelings and make normal moral judgments. That is, you would “see” the same good and evil as everyone else with the same innate moral standard. Good and evil would not be ontologically objective, but they would be subjectively the same for most people. Also, the compulsion to act toward moral good and away from evil would be instinctive, so moral value would also be subjectively normative (for people with normal moral instincts).

Moral value would not be objectively normative. It would just be in the mind of subjects with a certain type of brain. And if someone had different moral instincts, he would see good and evil differently, or not at all. There would be no way to persuade such a person to change his moral judgments, because there would be no higher standard or authority to appeal to. Imagine someone who simply lacks the moral sense. To him, nothing would be good or evil.

However, there doesn’t seem to be a moral sense. Humans have social instincts, and morality is partly a consequence of those instincts, but there is nothing like the moral realist’s conception of moral value in human instincts. Morality depends on the context: the culture, the society and even the individual. People do not “see” the same good and evil. This shows that morality is not simply the expression of a common moral sense.

In every war, each side views itself as morally good and the other side as evil. Both believe that they are acting toward moral goodness, even when they are massacring civilians. Acts that are considered the epitome of evil in the modern West, such as killing babies, torturing children and genocide, were considered good in other times and places, when directed toward enemies. Western societies have committed atrocities in recent history, such as the mass killing of civilians in WWII by the British and Americans. Somehow, people find ways to morally justify these actions when it is convenient, while condemning the atrocities of the other side.

Human history is extremely bloody. This poses a problem of evil for humanists, as it does for theists. If humans are innately empathetic and compassionate, why all the killing, torturing, enslavement, etc? But if human beings are innately selfish, it is easy to explain. Individuals compete for resources. Individuals can form cooperative groups (societies), but this just raises the competition to a higher level (war between societies).

The evidence of history is consistent with the theory that humans are selfish, not altruistic. This should not be surprising, because evolution creates selfish reproducing machines, not altruism machines.

We have the capacity for kindness, but we also have the capacity for indifference and cruelty. We evolved to form cooperative relationships, such as the pair bond and friendship. Such relationships are not altruistic. They are cooperative. Both sides benefit, and the relationship is conditional on reciprocity. Friendship and romantic love are both selfish, not altruistic. We also evolved to love our children, in a way that is mostly unconditional. That is energetically altruistic, but it is reproductively selfish. Creating descendants is the biological purpose of life. We also evolved the capacity to hate enemies: those who threaten us or compete with us.

Human nature is not altruistic. We evolved to cooperate, compete and care for offspring in ways that are instrumental to reproduction. Again, this is exactly what evolutionary theory predicts.

Most people who claim that we evolved to be altruistic are making some kind of group selection fallacy. They are falsely thinking of natural selection as selecting for what is “good for the group”, rather than what makes individuals reproduce. Even if we had evolved to act toward the good of the group (society, race or species), we would just be selfish at a higher level. The “self” would be the group. But that’s not how evolution works, or how we work.

Now, let’s consider the third option: that moral value is the cosmic telos, or instrumental to it.

This is analogous to the first one, with the cosmos replacing God, and the cosmic telos replacing God’s will. It has the same problems.

First, the moral realist would have the burden of proving that there is a cosmic telos, and that the human notion of moral value is instrumental to that telos.

To the extent that we can infer general principles of the cosmic order, the cosmos has no telos — or the telos is simply the conversion of ordered energy into chaotic energy, leading eventually to the heat death of the universe. That bleak view might be incorrect, but there is certainly no evidence for any other cosmic telos.

What if we look, not at the cosmos as a whole, but at life only? There is no telos to life as a whole. Evolution tends to generate complexity, so you could argue that complexity is the telos of evolution — but that’s not a very accurate way of understanding evolution. Evolutionary change is not in the direction of greatest complexity. It just tends to increase the complexity of life over time, while also limiting its complexity. Mutation adds complexity, but selection removes complexity. Most branches on the tree of life are pruned by evolution.

Nature is genocidal. Evolution proceeds by genocide. Types of life are created and destroyed, created and destroyed.

Each organism has its own telos: to reproduce. An organism is a reproducing machine. Its form was selected to have that effect. Natural selection is not based on aggregate or group outcomes. It is based on individual outcomes. Thus, biological purpose is at the level of the individual organism, not the species, and not life as a whole. Evolution creates selfish reproducing machines, not altruism machines or cosmic-value maximizing machines.

For more about evolution and biological purpose, see Debunking the Selfish Gene.

Do ecosystems reflect some natural telos? The same process that creates the tree of life also creates the order of ecosystems. Organisms reproduce to excess. There is variation among organisms. In an ecosystem, this variation includes the different species that occupy the ecosystem. Some organisms reproduce more than others. Most die young, without reproducing. This constant culling of the unfit (especially the young) creates the order of an ecosystem.

Nature is infanticidal.

The balance of nature is an equilibrium under selection, in which each species in the ecosystem just replaces its population, without increase or decrease. In this condition, most organisms of all species die young, without reproducing. A balanced ecosystem is not a utopia. It is a state in which the struggle for life is balanced. Life is so difficult that no species has an increasing population.

See The Balance of Nature.

Like the God of the Bible, nature is evil by modern standards. Nature is genocidal and infanticidal. Life is a brutal competition for finite resources. If life has a telos, it is nothing like modern moral intuitions.

But life, as a whole, has no telos. You, as an organism, have a telos: to reproduce.

Biological value is emergent from physical causality. It emerged via the loop of reproduction, by which organisms create new organisms. Thus, biological value is organism-relative. An organism has a telos. A species has no telos, and life has no telos.

See What is Value?.

A cosmic telos would be ontologically objective, and (in a way) objectively normative. But it would not be subjectively normative.

As with God, even if there was a cosmic or natural telos, I would be free to reject it. There is no logical implication from “X is the cosmic/natural telos” to “I should work toward X”. The is | ought gap lies between those statements.

Morality is a delusion, like belief in God. We are selfish reproducing machines. Our brains generate desires. We can form groups to cooperate. We can create collective values by social agreement and enforcement. Collective values are not cosmic. They are not imposed on us by logic. They are imposed on us by social incentives.

See What is Morality?.

There is no objective good or evil.