Friday, January 19, 2007

Beliefs and Moral Judgments

I have been accused (in the spirit of a civilized intellectual disagreement) of not giving proper weight to the importance of beliefs in behavior modification. I wish to answer that challenge by looking into the role of belief in ethics.

The Role of Belief

This blog is written on a foundation that uses the following formula to explain intentional action:

(Belief + desire) -> intention -> intentional action.

This equation shows that beliefs are a necessary part of intentional action. An agent that has desires but no beliefs will have no idea how to act so as to fulfill those desires. He will lay there and die without ever being able to answer the question, “What should I do?”

However, beliefs do not carry any moral weight. This is an ethics blog. As a result, the fact that beliefs carry no moral weight means I am not going to focus on them in defending moral conclusions.

The Cases

To see that moral evaluations track desires, rather than beliefs, consider the following set of cases.

Case 1: The woman desires to feed the infant. She believes that the bottle contains wholesome baby formula, and the bottle does contain good formula. She feeds the baby. The result is a nourished baby.

Case 2: The woman desires to feed the infant. She believes that the bottle contains good formula. However, the bottle contains poison. She gives the bottle to the baby, and the baby dies.

Case 3: The mother wants to feed the child. She believes that the formula has been poisoned. She is mistaken. However, because of her belief she does not feed the baby and it goes hungry.

Case 4: The mother wishes to feed the baby. She believes that the bottle has been poisoned. As it turns out, she was right. She refuses to feed the bottle to the baby. As a result, the baby is hungry, but alive.

Case 5: The mother wants to poison the child. She believes that the bottle contains wholesome baby formula and the bottle does, in fact, contain good formula. On the principle that a person must have a reason to perform an action, she does not feed the baby. The baby is hungry, but alive.

Case 6: The mother wants to poison the child. She believes that the bottle contains good formula, though it does not. Since the mother has no reason to give the formula to the baby, she does not do so. The baby is hungry, but alive.

Case 7: The mother wants to poison the child. She believes that the bottle contains poison, but she is mistaken. She feeds the formula to the baby. The result is a nourished baby.

Case 8: The mother wants to poison the child. The believes that the bottle contains poison, and the bottle does, in fact, contain poison. She feeds the contents of the bottle to the baby, and the baby dies.

Here, too, we can see how each of these cases follows the formula I mentioned above. Each time, the agent acts to fulfill her desires, given her beliefs. Beliefs end up playing a crucial role in what the agent does.

Now, let’s set Case 2 next to Case 8. Both cases involve a mother poisoning her baby. In Case 2, a mother was trying to feed her baby. However, she did not know that the bottle she thought contained food actually contained poison. In Case 8, a mother who wanted to poison her infant did so using a bottle that she knew contained poison.

For the sake of this demonstration, I want to eliminate certain pieces of clutter. Let us assume that the mother in Case 2 had every reason to believe that the bottle contained healthy food, and no reason to suspect poison. Assume that the poison was put into the formula from her husband or a disgruntled employee at the factory, where the mother had no way of knowing this.

The Irrelevance of Actions and Consequences:

In Case 2 and Case 8, the actions and the consequences of those actions are the same. In both cases, the mother feeds poison to her child. In both cases, the child dies. If we are going to base our judgment on actions or consequences, we would have to judge both mothers to be equally guilty, or equally innocent. It is absurd to hold that both mothers are equally guilty of murder. Thus, it is absurd to assume that moral evaluations track actions or consequences.

Note: If we include desires in our definition of an action, you get different results. But this will be shown to be consistent with the claim that moral evaluations track desires.

The Irrelevance of Beliefs

Does the difference in moral judgment rest on a difference in beliefs?

In this case, the mother in Case 2 had a false belief. She thought that the bottle contained healthy food, when in fact it contained poison. The mother in Case 8, on the other hand, knew that the bottle contained poison, and used it to kill her infant.

It would seem that if we are going to make moral judgments based on beliefs, that the mother in Case 2 who accidentally poisoned her child deserves the greater blame. She had a false belief. It would seem at least initially plausible that moral assessments grounded on belief would first track whether or not a belief was true, and give more moral credit to the person with true belief than the person with false belief. Yet, in this case, the mother with the true belief is a murderer, while the mother with the false belief was the victim of somebody else’s evil deeds.

Remember, we are assuming that the mother in Case 2 had no way of knowing that the bottle was filled with poison. We will deal with instances in which she should have known a little later.

Here, my interest is in showing that morality does not track beliefs. To see this further, let us assume that the mother in Case 2 acquires the same beliefs as the mother in Case 8. If morality tracks beliefs, we would expect to see the mother in Case 2 acquire the negative moral evaluation of the mother in Case 8. However, when we give the mother in Case 2 the belief that the bottle contains poison, we end up with Case 4 – a case in which the mother does not feed the child and the child lives. This case is not one of moral condemnation. Case 4 and Case 8 are not morally identical, even though their beliefs are identical.

Morality does not track beliefs.

The Morality of Intentions

Some people assert that ‘intentions’ are what matter. The mother in Case 2 did not intend to feed poison to the baby. She intended to feed the baby.

The problem with intentions is that they are ambiguous. The mother did not ‘intend’ to feed poison to the baby. Yet, at the moment she picks up the bottle, we can say that she intends to feed the contents of the bottle to the baby. It just so happens that the bottle contains poison.

Yet, even at the moment she picks up the bottle, it would not be accurate to say that she wants to feed the contents of the bottle to the baby. If another person, who knows that the bottle contains poison, were to see her picking up the bottle and say, “You do not want to do that,” he would be correct. She would ask “Why not?” He would answer, “Because that bottle contains poison.” Then she would realize that, in fact, she does not want to feed the contents of that bottle to the baby. She never did want to to that – though she did not realize it at the time. However, it is still true that she intended to feed the contents of the bottle to the baby.

Moral evaluations do not follow intentions.

Moral Evaluations Follow Desires

The only category that remains in this example is that of desires. Morality tracks desires.

Earlier, I mentioned how the mother in Case 2, if she acquired the same beliefs as the mother in Case 8, would not have fed the poison to the child and would be free from any type of moral condemnation. She does not acquire the moral status of the person in Case 8 simply by adopting her beliefs.

We can conduct the same test here, and give the mother in Case 8 the desires of the mother in Case 2. As the mother in Case 8 loses her desire to poison the baby and acquires a desire to feed the baby, she moves into the same situation as was described in Case 4. Case 4 involves a mother with a desire to feed the baby and a belief that the bottle contains poison. She refuses to feed the baby from that bottle, thus leaving the baby hungry and alive. Here, we see that the moral evaluation does track the desire in that, as the desire changes, so does the moral evaluation.

Yes, it is the case that the agent’s action also changes. This is because actions are the product of beliefs and desires. However, let us not forget that we began this investigation with two identical acts, and they showed that two mothers performing identical acts having identical consequences are not evaluated equally.

It is a fallacy to take somebody who has argued, “Not A” and answer, “I assert B.” Then, facing an argument that demonstrates “Not B”, assert “A” as if the first argument does not exist. The combination of the arguments I have given show that morality does not track actions, consequences, beliefs, or intentions. They track desires.

Crime Scene Investigation

Now, assume that you are a crime scene investigator. On this particular day, you have two infant deaths to investigate. In both cases, a mother has fed her infant a bottle that contained poison. The baby has died. Your job is to determine if a (moral) crime has taken place.

The first thing to note is that you have identical acts and identical consequences (for all practical purposes). Yet, you cannot even start to assign culpability in this case. There are an infinite number of possible stories to reveal. In some of them, the mother herself is as much a victim as her child. In others, the mother is the culprit. The important thing to note is that here, where you know the act and its consequences, you do not know enough to make a moral judgment. Moral judgments do not, in fact, attach to acts or consequences. This is not where the investigation ends, it is where the investigation begins.

In looking for culpability, what are you trying to find out?

In one case, you have your lab determine the type of poison involved and where it is available. Here, you discover that the mother had access to the poison. Furthermore, by looking at internet logs, you discover that the mother was researching the poison in the recent past. You also discover a credit card receipt (with her signature) where she bought the poison.

Through much of this, you are trying to discover what the mother believes. Mostly, you want to know if she believed that there was poison in the bottle when she fed the baby. You discover that she has beliefs about what the poison can do and how to get a hold of it. You also discover evidence the mother put the poison in the bottle while she prepared the formula. All of this leads you to the conclusion that she believed there was poison in the bottle when she fed it to the baby.

In all of this, you do not once blame the mother for her beliefs. In fact, you believe many of the same things she believed. You have the same beliefs about what the poison would do to the baby, and that the mother had mixed the poison with the formula before feeding the baby. If belief were reason for condemnation, you would be in nearly as much trouble as the mother.

So why are you interested in these beliefs?

Because you know the formula:

(belief + desire) -> intention -> intentional action.

You also know that people act so as to fulfill the more and the stronger of their desires, given their beliefs.

Without these assumptions, you would not be able to do your job.

If you take these equations and fill in the variables about what the agent believed, then you can then make inferences about what the agent desired. If you know that the agent believed that there was poison in the bottle at the time she fed the children, and she knew what the poison would do, you can deduce that she lacked any particularly strong aversion to bringing about those consequences. Ultimately, it is this fact that she did not care enough to preserve the baby’s life that determines her moral culpability.

What about excuses? Once you know that the mother has fed a bottle containing poison to her baby, what types of things can she use as an excuse to argue that she should not be blamed?

I have a section devoted to the subject of excuse in my book. One of those possible excuses turns out to be a ‘mistake of fact’. That is to say, “I did not know that the bottle contained poison.”

Assume that, in the second case, you check the formula and discover a manufacturing lot number. Checking the lot number, you find other packages also containing poison. The mother had no opportunity to deliver poison to the factory, so somebody else must have done it. In all likelihood, the mother had no way of knowing about the poison. Given her beliefs, even a mother who cared deeply about the welfare of her child would have fed the formula to her baby. There is no reason to believe that this mother lacks the desires of a good mother, so no reason to condemn her. In fact, we have every reason to see her as a victim.

There are several forms of excuse – from ‘consent’ to ‘accident’ to ‘greater good’. What all of them have in common is the fact that they block any inference from our knowledge that a particular act took place to the conclusion that the agent had bad desires or lacked good desires. All excuses make it plausible that an agent with good desires would have performed the same action.

Epistemic Negligence

There is one set of cases where it may seem that we are holding people responsible for their beliefs. This is when a person believes something absurd and, as a result of this absurd belief, acts in ways that harm others. In extreme cases, this may be considered insanity. In other cases it is a more common form of negligence. An example of the latter is a case where hunter fires at what he thinks is a game animal, only to discover that he has shot his friend. He negligently formed the belief that the creature making the noise in the brush was a game animal.

Epistemic negligence is very much like physical negligence. Compare this to a case in which a rancher overloads his pickup with hay bails. On the drive home, some hay bails fall off, land on another car, causing an automobile accident that kills two family members. The farmer in this case is guilty of negligence.

There is no reason to view epistemic negligence any different from physical negligence. The epistemically negligent is somebody who fails to properly secure a dangerous belief, in the same way as the farmer failed to secure a dangerous load.

Cases of physical negligence are cases of a defect in desire. We can infer from his lack of care to protect the welfare of others that he lacked the concern for the welfare of others that we all have reason to promote in society. We condemn the negligent for this lack of desire.

It is the very same lack of desire that causes the epistemically negligent to secure their beliefs. If they were truly concerned about innocent people being harmed, then they will take care to make sure that beliefs that might cause harm are properly anchored and secure. If they fail to take the necessary precautions to secure these beliefs, we may infer that they do not care about the harm they do to others.

The best part of the newest rise in atheist activism is that atheists are taking epistemically negligent beliefs that are a danger to others and giving them the level of moral condemnation they deserve. These are religious beliefs that support the conclusion that others are to be harmed. Beliefs that others are to be harmed deserve a far more secure foundation than religion can possibly provide. Those who do not go to the pains to make sure that their beliefs that others are to be harmed are well secured may be morally condemned as people who lack sufficient care that they not inflict unnecessary and unjustified harm on others.

Bigotry

Finally, we have the fact that people have a tendency to believe what they want to believe. From this simple fact we can sometimes make inferences about what another person desires simply by looking at his beliefs.

It is important to note that this type of inference only works where an individual makes some sort of mistake. Either she bases her conclusion on evidence that is not well supported, or she makes some flawed inference. When people make a mistake like this, we have reason to ask, “Of all of the mistakes that you could have possibly made, why did you make that mistake and not some other?” The reason will often be because that mistake fulfills some desire of hers.

For example, let us take the bigot who believes that hurricanes hitting New Orleans has something to do with the town’s greater-than-normal acceptance of homosexuality. This is clearly an absurd belief. We know enough about hurricanes to know that our computer simulations that predict their paths do not need to include a variable for “coastal region’s acceptance of homosexuality.”

When people make a mistake like this, we then have reason to ask, “Why did you make that mistake? Why did you accept that absurdity?” The answer, in this case, is that they are motivated by hatred, and are looking for something – anything – that might give their hatred legitimacy. There is no good reason to hate homosexuals. Therefore, these agents level completely unfounded and unjust charges against the victims of their hate. In this case, it is the charge that homosexuals attract hurricanes.

In this case, from a mistaken belief, we can infer something about the agent’s desires. From what we then know about the agent’s desires, we can make a moral evaluation. In this case, we know that the agent is a hate-mongering bigot unfit for civilized society.

Conclusion

So, what can we say about the role of belief in ethics?

Beliefs clearly play a role in influencing intentional action. However, beliefs are morally inert. In fact, actions, consequences, and intentions also fail to properly explain and predict moral judgments. The only claim that actually appears to work is the claim that moral evaluations track desires. The moral value of a person’s actions comes from the desires she has, or the desires she does not have that a person with good desires would have.

Criminal investigations can be best understood as a scientific quest to discover the desires of the agent that motivated a particular act. The act and its consequences alone are not enough to fix culpability. Beliefs are useful, but they are not the reason for blame. What matters is what the accused wanted to do with those beliefs. It is her desires that determine culpability, not her beliefs.

There is moral culpability for negligent beliefs – beliefs that are a danger to others that an agent fails to tie down securely. This need for well secured beliefs is why we have reason to require federal agents to go before a judge before they spy on or arrest people and why we have a right to a trial by jury. However, this is not because of the moral quality of the belief. Epistemic negligence is a moral crime because it demonstrates that the agent really does not care about the harm that her unsecured beliefs may due to others. It is this lack of compassion that deserves our moral condemnation, not the false beliefs.

Also, whenever a person clearly adopts a belief that is clearly absurd and grounded on no solid evidence, we can use this to infer what the agent wants. People have a tendency to believe what they want to believe, so any beliefs that are not well secured will tell us what the agent wants to believe. The bigot who makes up reasons to hate Jews, gays, or atheists can be shown to people who embrace hatred. If they were truly opposed to hatred, they would see the flaws in these types of arguments.

This, then, is a quite insufficiently brief outline of the role of belief in morality.

6 comments:

Anonymous said...

Alonzo --

I have said you don't give beliefs enough weight, but I haven't said beliefs should be weighed in moral judgements.

I've argued that it is in considering what to do about immoral behavior or how to prevent immoral behavior that you do not give enough weight to beliefs. You tout condemnation and praise while seeming mostly dimissive of addressing beliefs.

As I stated previously, in the real world where trying to choose a course of action based upon morality can be complex and nuanced, I would not want to be surrounded by people who have acquired there morality through the conditioning you propose.

Thayne

Alonzo Fyfe said...

Thayne

There is no other way to acquire morality.

If moral judgment tracks desires, and desires are not 'true' or 'false', then we cannot affect a person's morality by affecting what they hold to be 'true' or 'false'.

The only way to affect a person's morality is by affecting what they like and dislike.

I have argued that we have reason to be concerned about beliefs - particularly about epistemic negligence. However,even when it comes to morality of beliefs, the best we can do is to give the person a love for well-anchored beliefs when those beliefs threaten harm to others.

Not liking a particular conclusion and wishing things were different is not an argument that it is different.

Alonzo Fyfe said...

Look to tomorrow's post for an explanation as to why it is not possible to reason people into being moral.

We can reason people into believing what the moral thing to do is. However, knowing that something is right, and doing the right thing, do not coincide.

Anonymous said...

Alonzo --

Except that most people have some desire to be good. By making a strong case that what they are engaged in is immoral, many people will change their behavior, and perhaps even re-evaluate why they do the things they do. Some people just don't think much about the implications of their behavior -- and that can be changed through reason. This is not some armchair theory of mine, it is something I, and I'm sure many others, have seen.

Furthermore, wouldn't conditioning be much more effective if combined with demonstrating why a thing is wrong or right?

And, how will your reasoned (well reasoned, I would say)blogging on moral theory make the world a better place if it cannot possibly make anyone moral? I'm not saying there isn't an answer to that, I just wonder what yours is.

Thayne

Anonymous said...

Alonzo --

I should actually point out that I do agree with you in that reasoning cannot make someone want to be moral. A person who truly and strongly desires to harm others will not be persuaded by reason.

But I think few people are like that. Most do bad things because they are able to convince themselves that what they are doing is actually okay, or they simply blind themselves to the consequences of their actions. For such people (most people), reasoning, I contend, can and should be an important part of changing their ways.

Thayne

Alonzo Fyfe said...

Thayne

Yes, many people have a desire to be good people, or a desire to do the right thing.

A person with a desire to do the right thing, and a belief that X is the right thing to do, then that person will be motivated to do X.

Yet, we still have a question to answer:

What does it take for a belief that X is the right thing to do to be true? What theory best accounts for the rightness of doing X? We still need an answer to this question.

You will find some of the answers to your comments here in tommorow's blog entry. There, I argue that reason has a role to play identifying the desires that there are more and stronger reasons to promote or to inhibit. Yet, reason alone cannot promote or inhibit any desire.

An cigarette smoker can be convinced that he would be better off without a desire to smoke. However, no amount of reason can change the fact that she has the desire to smoke. To actually deal with this desire, she has a lot of other work to do.

Reason can tell you the best way to get a job done. Reason alone cannot do the job.

I found the following comment interesting:

And, how will your reasoned . . . blogging on moral theory make the world a better place if it cannot possibly make anyone moral?

Well, how can well reasoned medical research make patients healthier if it cannot possibly make anyone healthy? Well reasoned medical research makes nobody healthy. However, it does tell people how to go about making others healthy. My blogging may not make anybody moral. However, it does (I hope) give an idea of how to go about making others moral.