On causality

This is some very old stuff, in which – with truly breathtaking hubris – I attempt to improve on Aristotle. Professional philosophers will derive a great deal of innocent amusement from it, which amusement I do not begrudge them in the slightest. If anybody else finds something useful in it, you’re welcome to it.

I’m resurrecting it as a sidetrack off of a discussion in the comments of All Things Beautiful. Michael, the particular sections you would be interested in are the introduction and the section on volitional causality.

If I were writing this today I would work a lot harder to make it readable, but I’m not going to go to the trouble of a rewrite at this point.

...continue reading...

IV. Mistakes in Reasoning (the Emotional Fallacies)
continued

C. False causes: clarifying the sources of problems and solutions

Why did this happen? How can I make this happen? Who is responsible for something that has happened (either bad or good)? Whenever we ask one of these three questions, we enter the realm of causality. You cannot solve problems without addressing cause and effect; you can hardly make decisions without taking into account the effects of which your decision could be the cause; and certainly you cannot attach blame to a person without implying that that person's behavior or negligence caused the event for which you blame him. Consequences and results, problems and solutions, means and ends, responsibility and reward -- none of these issues can be addressed successfully if we are not capable of accurately weighing causes and effects.

Causality is, however, a nastily complicated subject, the kind of subject philosophers just love. I'll try to simplify as much as I can, but it will still be difficult, if not impossible, to avoid introducing a number of technical terms. I'll minimize the technicalities as much as possible.

There are two basic forms of the fallacy of false cause. One is equivocation: just as analogies give their own peculiar twist to overliteralism, so causality gives its personal stamp to equivocation. The other is the false cause properly so called. We'll deal with the equivocation first, and then turn to the various ways people fall into false cause fallacies.

1. Equivocation

There are a number of terms that need to be very carefully defined. We need to be able to work out the differences between causes and factors, between main causes and secondary causes. And we need to examine the four or five different kinds of causes. Traditionally these have been Aristotle's tetrad of the efficient, final, formal and material causes, but, for the sake of those of us who are not into formal philosophy, I have instead devised my own division into material, situational, formal, motivational, and volitional causes, a division which is more closely lined up with the categories of modern thought and therefore easier for the average American to understand. (This, by the way, is the only superiority I claim for my scheme of distinguishing between kinds of causes.)

a. Primary causes and secondary factors

i. Causes versus factors

First, what is the difference between causes and factors? Let me explain by way of example. If I punch you unexpectedly in the face, then the reason your nose is broken is that I punched you in the face. The cause is my punch; the effect is your broken nose. Clear enough? But now suppose that I am talking to you, and you are listening: in short, we are communicating. Now, what is the cause of our communica­tion? Is it my talking, or is it your listening? Which is it? Well, of course it's both; neither would do any good without the other. Now, in this situation we could say that there are two causes. But for the sake of clarity, in our discussions I would prefer to say that there are two equally important factors involved: my talking, and your listening.

Now in ordinary English, we tend to use the word cause to refer to the one factor that we consider most important. For example, I think everyone would agree that if a bystander had not videotaped the Rodney King beatings, riots would not have broken out in L.A. upon the news of the jury's verdict -- for it was the incessant replaying of that videotape that caught the attention, and ignited the indignation, of the nation. So the videotape could technically be called a necessary cause of the riots: if the videotape had not been made, the riots would not have taken place. But would anyone say that the videotape was "the cause" of the riots? Put it another way: if we passed a law stating that it would henceforth be illegal to videotape police brutality, would we feel that we would henceforth be safe from the threat of race riots? Obviously not. No, it would be better to say that the videotape was a factor -- and a necessary factor -- in the process which culminated in the riots, but it was not the cause, or the main cause, if we wish. It was at best only a factor, or a secondary cause.

I propose to use the terms main cause, main factor, primary cause and primary factor inter­change­ably, as referring to the most important cause of a given effect. The term factor will refer to any cause which contributes to an effect, but which is not necessarily the main cause. And the terms secondary cause and secondary factor will refer to factors that are known not to be the main cause.

With those definitions, it becomes obvious that a common form of equivocation will arise when someone proves that some secondary factor (e.g., the videotape) is a "cause" -- in the sense of "secondary factor" ‑‑ of some effect (e.g., the riots), and then from there proceeds as if said secondary factor were in fact the primary "cause" of the effect (e.g., "...so if we outlaw private videotapings of police activities, we will ensure peace and order in our inner cities").

Now the alert student will have noticed that the term "most important cause" is, as it stands, a rather subjective one. "Most important from what point of view?" this student asks. "Is there any objective way to measure causal importance?" Well, there are some ways to lessen the subjectivity. We should distinguish between near and far causes (what classically are known as proximate and ultimate causes respectively). And we want to have a handle on the concepts of necessity and sufficiency, and on the distinction between root causes and surface causes. (Notice how the terms are multiplying? This is what I meant when I said that the fallacy of false cause demands more philosophy from us than any other fallacy.) On the other hand, the question of which type of cause is the most important is usually subjective: we don't know until we find out from what point of view we're approaching the issue.

ii. "Most important factor" -- near versus far (proximate versus ultimate)

Let us say that I live in a house in the forest, surrounded on all sides by small meadows of clover. Then lightening strikes a tree several miles away. A forest fire results. The wind blows the fire toward my house. When the fire reaches the clover fields, the grass catches fire, which in turn allows the fire to cross the meadow and reach my house. My house then burns down. Now, what is the cause of my house burning down? Is it the lightening striking the tree? Is it the direction of the wind? Is it the fact that the fire crossed the clover fields to reach my house? Well, it's all of them. But the fact that the fire crossed the clover is the nearest (or proximate) cause; the wind blowing the fire my direction is sort of in-between, and the lightening striking the tree is a relatively far (ultimate) cause. If I get it into my head that the main cause is the lightening, then I am liable to say, "Well, there's nothing I can do about lightening, so I couldn't have saved my house." But of course this is not true; if I had created a firebreak at the clover fields, the chain of causality would have been broken at the nearest cause, and my house would have been saved.

iii. "Most important factor" -- sufficiency

Now let's introduce the concept of sufficiency. A sufficient cause is one that doesn't need any help to produce its effect. If you have a sufficient cause, you'll get the effect; there's no way out of it. Perhaps the easiest non-technical way to look at it is this: whatever takes you past the point of no return is a sufficient cause.

Look back at my example of the house. If I'm there, and the fire's headed my direction, I can still save my house by setting up a firebreak. So the lightening isn't a sufficient cause; my house could survive despite the lightening. But if I don't get the firebreak made, then once the fire gets to my house, my house is a goner (assuming that I'm out in the country and don't have a fire department or water handy). The point of no return is the point at which the flames actually reach the house, so the farthest we can go back and still find a sufficient cause is at the point where the fire crossed the field. However, if I'm not home, then all that's necessary is for the wind to be blowing the right direction, because if it is, nothing can stop the fire (there being nobody there to make a firebreak and thus nothing to keep the fire from crossing the field). So in that situation the wind blowing the fire toward my house becomes sufficient. If I live in an area where a consistent prevailing wind is always blowing in the direction of my house, and I'm not home, then the lightening starting the fire is sufficient: once that happens, the house is doomed. The principle, you see, is that once you find a place where the chain could have been broken, all the causes farther back than that are insufficient.

Sufficient causes are always more important, ceteris paribus, than insufficient causes.

We should note, however, that whether a cause is sufficient or not depends to some extent on what we take as "given," as an unchangeable factor in the equation. For example, the primary cause of any riot is the people who choose to throw the riot. Politicians, however, like to talk about how the riot was "caused" by their opponents' policies, or by some particular event that they can blame on their opponents. In doing so, they are implying that the alleged "cause" was a sufficient cause, that it took us past the point of no return. And they can only do so, logically speaking, by taking it as an unavoidable fact that people will choose to riot, given a good enough reason to do so (which the event is presumed to have been). If we accept that statement as true, then we can call the triggering event a sufficient cause. But in fact people to react to triggering events in different ways: some riot, others content themselves with refusing to vote in the next election. Any riot could be avoided, no matter what the provocation, if the rioters would simply choose not to riot. ("What if they had a riot and nobody came?") So strictly speaking, no trigger event can be sufficient cause for a riot.

Yet for practical purposes we may have to treat it as such. Look at a slightly different example. I know that the school bully is mad at me and wants to beat me up. Now, if he would just change his mind about beating me up, I would be safe. So walking past the place where he hangs out would not be sufficient cause for me to get beat up, in the strict sense. But the fact is that I can't make him change his mind about beating me up, and it seems very unlikely that he will do so on his own. Meanwhile I am quite capable of avoiding the pool hall where he plays video games. So the wise course is for me to take his belligerence as an unchangeable fact of nature and behave as if all it takes for me to get beat up is for me to get within reach of him. From my perspective, I should proceed as if walking past his hangout would constitute a sufficient cause of pain and suffering.

iv. "Most important cause" -- necessity

Necessity is the third standard of importance for causes. A necessary factor is one without which the effect could not take place. Turning once again to the L.A. riots, let us try to decide whether the basic cause was the criminal character of the rioters or whether it was the King beating and verdict. Neither the criminality nor the verdict is sufficient cause, for if nobody chose to riot, there would be no riot, yet riots have to have a trigger event in order to get a large enough number of willing rioters all mad at the same time. So we ask which is a necessary cause. Do you have to have rioters in order to have a riot? Absolutely. It is absolutely necessary. Do you have to have an unjust verdict in order to get a riot? Nope. In the past three years, we have seen full-scale riots thrown in two separate cities (Detroit and Chicago), in each of which cases the trigger event was the local NBA team's winning the NBA championship! When this fact is taken into consideration, it becomes obvious that just about any excuse will do: our inner cities have a lot of people in them who enjoy destroying and stealing things and who find random destruction and looting to be an acceptable way of expressing strong emotion, and if you get them stirred up, you'll find yourself having to deal with a riot. (This assertion is born out, I think, by the fact that a large proportion of the L.A. rioters were actually Hispanic rather than Black ‑‑ "Hey, guys, check it out -- a riot! I can get into this! Gimme that gasoline!" ‑‑ and by the fact that the businesses that suffered were in many cases owned by Blacks.)

Now, does this mean that we shouldn't care about racism in the legal system? Of course not. Racist discrimination in the courts is a betrayal of everything our legal system stands for, and it cannot be tolerated. Does it mean that racism does not result in riots? No, it doesn't even mean that. For racism may well play an important role in turning inner city kids into criminals. All that our discussion has shown is that reforming the legal system will not in itself prevent riots in the inner city. If we want to solve the problem of riots in the inner city, we must attack the problem of the criminal character of so many inner city kids and adults. The root problem is that there are a lot of warped people in the inner cities. Solve that problem, and you will have solved the problem of inner city riots. But if you try to attack the problem solely by going after trigger events, then what are you going to do -- outlaw NBA championships?

So a necessary factor is, ceteris paribus, more important than a non-necessary factor, just as a sufficient cause is, ceteris paribus, more important than an insufficient factor.

We can sum up by saying that the ideal cause would be one that was both necessary and sufficient. It is the relationship described in mathematics by the phrase, "if and only if." The factor that comes closest to being the ideal cause, i.e., necessary and sufficient, is usually the "most important factor."

v. "Most important cause" -- root causes vs. surface causes

Finally, we look at the concept of the root cause. The root cause is always farther back than the surface cause. Root causes come into play when we consider solving problems. What happens is this: you have a far cause that is sufficient to produce a certain class of effects. But the form taken by the effect is determined by situational causes. Since we don't yet know what situational causes are, I think it best to put off the discussion of root causes vs. surface causes until we have gone over situational causes. In the meantime, the main thing to keep in mind about root causes vs. surface causes is simply this: if you try to solve a problem by attacking the surface cause instead of the root cause, the problem just pops up again in a different form. E.g., the root cause of riots is rioters' love of violence; if you clean up the court system without trying to deal with rioters' love of violence, you ought not be surprised to see a riot occur because the Pistons win the championship, or because Bill Clinton is or is not elected President, etc. For inequity in the court system is just a surface cause that determines the time and place in which rioters gratify their love of violence.

Clearly, then, a root cause (a) is a relatively far cause and (b) is sufficient. Generally speaking, it is also (c) necessary. We'll come back to the term after we're comfortable with situational causes.
b. Kinds of causes: material, situational, formal, motivational, and volitional causes

Nearly twenty-five hundred years ago, Aristotle sorted causes out into four kinds: the efficient, the final, the formal, and the material. I, on my own authority, am tossing out Aristotle's classification in favor of my own, for one reason: I think mine are easier for the typical American to understand. But in other respects Aristotle's are probably better, and I refer you to Mortimer J. Adler's Aristotle for Everybody in hopes that you will at some point take the trouble to learn the classical scheme.

Basically, here are some very rough definitions:

Volitional cause: The person who chooses to do something; the exercise of free will (i.e., volition). I will generally use the terms choice or chooser to refer to the volitional cause.

Motivational cause: The motivation a person has for doing something. I will generally use the term motive rather than the longer, more technical term motivational cause.

Material cause: (a) The tools used by the person doing something. (b) The action of natural law.

Formal cause: When the best explanation for an effect is the essential nature (the Platonesque form) of something.

Situational cause: The conditions that make something possible. I will generally just call this the situation, or maybe the conditions.

Take a murder in a mystery novel. What causes the murder? Well, that depends on what you mean. You could answer, "Well, it would hardly be a murder mystery without a murder, now, would it?" In which case you are citing a formal cause: the form of a murder mystery requires that a murder be part of the plot. Or you could adopt the viewpoint of the master detective within the novel, and point out the murderer (volitional cause), or the motive (motivational cause), or the method (material cause), or the opportunity (situational cause). And all of these would be reasonable answers to the question, "Why did Mr. Q. die?" "Because the butler killed him," or, "Because the butler was afraid he was going to be fired," or, "Because every murder mystery needs a murder," or, "Because he was hit on the head with the silver chafing dish given to him by Lady Packard," or, "Because nobody except Mr. Q. realized that the butler was really Leon Acapulco, retired second-story man." Each of these is a perfectly good answer. And each interprets in a different fashion the ambiguous statement, "Why did Mr. Q. die?"

Please note, in passing, that while I have used the Aristotelian terms material cause and formal cause in my scheme, they are not used in exactly the way Aristotle uses them. They are reasonably close, especially if you take "the action of natural law" out of the material class and call it an efficient cause, but if you want to use Aristotle's terms with Aristotle's meanings, you need to study Aristotle.

i. Material causes

Material causes are the special province of science (which concerns itself with natural laws, i.e. with what happens in nature, given certain circumstances, as long as some intelligent being doesn't step in and interfere) and technology (which concerns itself with applying natural law in order to develop more efficient and powerful tools).[1] The scientific method was developed in order to exclude all nonmaterial causes from an experiment, and so to isolate material causes. For this purpose, there is no better method in the world.

But note that science, if applied to the investigation of non-material causes, is out of its league. Its methods are designed for searching out material causes; they cannot deal with motivations or choices (motivational or volitional causes). In plain English, if you want to know whether a girl loves you, you don't set up a double-blind lab experiment to find out. And if she agrees to marry you, but you aren't sure whether it's for you or for your money, the most precise knowledge of natural law will not resolve your doubts.[2]

Perhaps one of the most widespread sources of confusion in our thought comes from the combination of (a) the fact that the scientific method has proved magnificently useful in finding out material causes and (b) the fact that people use the term unscientific equivocally. In one sense, it is quite true that if someone asks, "How did life arise on earth?" and you respond, "God created it," your answer is unscientific. That is, you have cited a volitional cause rather than a material cause, and science can't deal with volitional causes. But so many people take unscientific to mean anti-scientific (as if your "unscientific" answer was the equivalent of denying the existence of dinosaurs by saying that God put fake bones in the earth in order to fool arrogant paleontologists) that religion is seen to be opposed to science. This error is compounded by the fact that there are lots of scientists who are philosophical materialists (i.e., who believe that all causes eventually reduce to material causes) and lots of religious people who distrust and revile the scientific community. But if a scientist claims that science disproves God's existence, he is laboring under the delusion that science deals with nonmaterial causes, while a Christian who claims that Christianity tells us what scientific beliefs to hold has probably confused the motivational and volitional concerns of religion with the material concerns of science.

Three other points to make about material causality:

(1) Science can shed light on volitional or motivational causes, but only indirectly. For example, an inquest jury is trying to decide whether a death was accidental or not. The autopsy shows that death was the result of the ingestion of granular arsenic; furthermore, science tells us that arsenic is not naturally found in any of the food or drink that the dead person ate just before he died. The jury may therefore be certain that death was not accidental, i.e., that it had a volitional cause (somebody made him die on purpose). Then the jury can turn to the question of whether it was suicide or murder -- a question which cannot be settled by a lab experiment, since it will turn on motive (motivational factors) and opportunity (situational factors). If the autopsy had, on the other hand, shown that death was due to a heart attack, it would have provided a natural (i.e., purely material) explanation of the death, but that would not prove that death was in fact accidental, since every detective writer who ever lived has written stories in which murderers deliberately frighten their victims to death. You would still have to consider the possibility of other causes, and science could not rule on them without showing that there are no tools available for inducing the particular cause of death. If it could show that no such tools were available, however -- for example, if death were the result of a cerebral aneurysm, and current technology could not induce a cerebral aneurysm -- then you could rule out murder, because the murderer (the hypothetical volitional cause) wouldn't have had access to tools (material factors) suitable for the job.

(2) Any valid science confines itself to material causes, even it does not appear to (either this, or it is not a "science" in the modern sense). Take psychology, which seems to be concerned with human beings, i.e. with volition. In fact, psychology (insofar as it is in the strict sense a science[3]) is concerned with humans as they are affected by the material factors of their subconscious, i.e. as humans who do not have free will and therefore are acting involuntarily. An example[4] will help make this clear: take two soldiers who are called upon to die for their country. One, a psychologically normal human being, makes the conscious decision to sacrifice himself, and does so, dying bravely. The other tries to force himself to do his duty, but he is crippled by a pathological terror of death, and finds himself fleeing in terror despite all his efforts to force himself to be brave. Now, let us suppose that a psychotherapist takes our pathological coward in hand, and cures him of his pathological condition. He is now a psychologically normal man. But this does not mean he will choose to act bravely. If he is psychologically capable of facing death, but is selfish and calculating, he will still choose to flee rather than to die. And that is a choice he freely makes, and psychology can't do anything about it, unless a propagandist uses psychology to manipulate the man, i.e., to steal his free will away again. So psychology and brainwashing are flip sides of the same coin: psychology tries to free a human from psychological (material) handicaps so that he can be a fully human, volitional agent, while brainwashing tries to enslave the free human will to psychological constraints and thus reduce him to a mere tool, a mere material agent. When a human being begins to act by choice instead of by psychological necessity, psychology cannot help us, for the subject matter of psychology is the human psyche as material, not volitional, agent.[5] That is why it can make sound use of the scientific method. And the same will hold true for any valid science.

(3) As was implied in the previous paragraph, human beings can be material causes. It happens every time one person uses another person as a tool. To use another person to accomplish your own ends is to make them a material factor. If you do so against their will and without their consent, you reduce them to merely material factors, which is a thoroughly dehumanizing way to deal with people.

ii. Situational causes

Situational causes are in one sense not really causes at all. They never actively make anything happen; when they influence events, it is by preventing something from happening or by making it possible for something to happen. Traditionally, situational causes have been labeled conditions, and it has been considered a mistake to treat them as a cause; thus you will read of the "fallacy of confusion of condition and cause." I certainly do not deny the existence of the fallacy; it's just that I think it more useful to call it a confusion of one kind of cause with another. For conditions do shape events, and in that sense are part of the chain of causality.

The distinction between a material factor and a situational factor is really rather simple. The material cause is generally positively active; the situational cause is negative. Well, that didn't sound simple, did it? We'd better get to an example. When I talk, and you listen, you will hear me, right? Well, not if a jet plane is taking off twenty yards away. In other words, it has to be quiet for you to hear me. The quiet is a necessary condition, i.e., it is a necessary situational factor. But is it sufficient? Of course not. We can be alone in a public library and you still won't hear me talk -- if I refuse to say anything, or if you are deaf. In other words, overpowering noise can force you not to hear me, but its absence can't force you to hear me. And whenever you find a situational factor that seems to be sufficient, i.e., seems to be an active cause, then one of two things is true: either the situational factor really just makes it possible for something to happen (as quiet makes it possible to hear, and as light makes it possible to see), or else you're calling something a situational factor when it's really a material factor.[6]

Let's take a more detailed look at the way situational factors take effect. Let us say that we have concluded that crime is a more severe problem in inner cities than in rural areas, and we have further concluded that this is because there is more poverty in the inner cities. In other words, we think poverty is a situational cause of crime. But how exactly does poverty cause crime? It isn't sufficient, for there are plenty of poor people whose integrity prevents them from resorting to crime, and it isn't necessary, for there are plenty of rich people who commit crimes right and left. Yet clearly it contributes to the problem.

Well, the first step in the analysis is to recognize that poverty is at most a situational cause, and that means there has to be some other cause as well. Since crime is committed by people who choose to commit it, there must be a chooser (a volitional cause), namely the criminal. And why does the criminal choose to commit his crime? That brings us to the motivational level. Let us say that we conclude that the motivational cause of crime is the greed and selfishness, and maybe envy and bitterness, of the criminal. In other words, the cause of crime is a corrupted human being trying to gratify depraved desires: in a word, a criminal.

Let us take, as an example, a hypothetical rapist with a perverted desire to dominate and tyrannize other people, and let us say that he has decided to find some way to gratify that desire. (Thus the motivational and volitional factors of desire and choice are present.) Rape is a form of domination, so it would satisfy the formal requirements, and we will say he is a normal, physically healthy and strong man, so that all necessary material require­ments are satisfied. But rape is not the only way to dominate people. You can go into athletics and become a Bill Lambier or a Mike Tyson; you can amuse yourself with hostile takeovers; you can be the leader of a group of kids who have fun ridiculing your school's nerds; you can become a demagogue wielding political power to break your enemies, etc. The problem is that to dominate people in athletics, you have to be a good athlete; to lead hostile takeovers, you have to be rich; to lead other kids, you have to be popular; to be a successful politician, you have to be reasonably personable and persuasive, as well as well funded. In other words, there are lots of lots of legal ways to dominate and hurt people, but they all require certain situational factors. So poverty generates its effects by denying this hate-filled person all the legal ways of satisfying his craving for domination, thus taking a person who might have become a rich, ruthless, but respected public figure had poverty not restricted his options, and "turning him into a rapist" -- since most men have enough talent and opportunity to commit rape if they set their minds to it.

Now that we know how situational factors come into play, we can consider how to make use of them. Basically, situational factors are the issue whenever we discuss risk. It is much easier for lightening to strike you if you're standing on top of a mountain than if you're sitting in a cave ten yards back from the opening. So if you don't want to get hit by lightening, don't stand on top of mountains during thunder­storms. One nearly infallible way to gauge wisdom and maturity is to see whether a person does a good job of controlling the situations he gets into -- because foreseeing the effects of situational factors demands imagination and foresight, neither of which are particularly characteristic of fools and teenagers.

I think the best way to illustrate the proper use of situational factors is to look at Billy Graham's example. Here is a man who for forty years has never been alone, even for five seconds, with any woman except his wife, and who never enters his hotel room without first sending a friend in to make sure no unpleasant surprises are waiting. Does the fact that he has never committed adultery mean that he is moral? Not at all. We can't know what his thought life is like. But by controlling his situation, he has made it impossible for himself ever to commit physical adultery, no matter what kind of temptation he undergoes. And he has furthermore made it impossible for anyone to slander him effectively. Most people would probably think the bit about sending someone into the hotel room to check it was pretty paranoid -- until they learned that once, when his advance scout entered the room, a nude young woman leaped out from behind the door, threw her arms around the scout (clearly she assumed he was Dr. Graham), and kissed him soundly, to the accompaniment of the flash of the camera of the photographer who had set up the ambush. One need merely imagine the bids the tabloids would have made for those photographs had Dr. Graham been first into the room, or think of the damage done to Jimmy Swaggart by his "making provision for the flesh [i.e., not controlling situational factors wisely]," to appreciate the difference situational wisdom has made to Dr. Graham's ministry and reputation.

But again, I must emphasize: situational factors do not cause something to happen. They make it possible for something to happen, or else they keep something from happening. A high school girl who winds up alone and drunk in a frat brother's bedroom while wearing a miniskirt, wet T-shirt and no bra has gotten herself into a bad situation. But if the frat brother proceeds to take advantage of the situation, she has not "gotten herself raped." The volitional cause of her rape (the one who chooses to make it happen) is her rapist. She has been foolish. But she has not "asked for it;" she is only a material cause in the rape (i.e., something used by her rapist), and her foolishness is merely the cause of a situational factor that contributed to the rape that her rapist used her to commit. To blame her for her rape as if she in some way consented to it or cooperated in it, is to confuse situational factors with volitional factors.

We now, I think, can see how root causes and surface causes interact. The root cause is a far material, motivational or volitional cause sufficient to produce a certain class of effects. Nearer situational causes determine which particular effect among that class gets produced. Thus the rapist's sexual lust, desire for dominance, and disrespect for the rights and dignity of others (the motivational causes) coupled with his willingness to gratify his desires if given an opportunity (the volitional cause) constitutes a root cause that is sufficient to produce any of a number of effects -- the purchase of pornography, patronization of prostitutes, sexual harassment, or rape, to name a few -- depending upon what opportunities he has to carry any of them out. Thus when he discovers himself in a situation where he can rape an attractive young girl, he rapes her. If she refuses to go to his room, he'll probably just look around until he finds someone more gullible, and rape her instead. All the girl can do is keep from being the one he rapes. Similarly, Dr. Graham's refusal to be alone with any woman besides his wife doesn't do anything about the root problem of sexual lust; if he assumes that he has taken care of that simply by avoiding adultery, he is likely to find himself trapped in some other problem (a distractingly vivid and immoral fantasy life, for example). All that he accomplishes by controlling his situation is to limit the particular way in which his sexual lust (I presume he has to deal with some) gets expressed.

For a political example of the consequences of confusing root causes with surface (i.e., situational) causes, consider the people who want to nationalize the health care system, or rather those who want to nationalize it in order to "assure adequate and affordable health care for every Texan."[7] The problem they perceive is that the average uninsured middle-class American can't afford medical care, because he doesn't have enough money. So they propose to fix it by having the government make sure everyone has plenty of money. Unfortunately, this ignores the root cause of the health care crisis, which is that there are too many people who want to be cared for and not enough people to care for them. Given a situation where insurance companies pay as much as is required for the insured middle-class, and government subsidies pay as much as is required for the qualifying poor, the people who get left out will be the uninsured middle class. But subsidizing them, while it fixes the surface problem (lack of money for health care) will not fix the root cause (shortage of medical care). You will still have to decide who gets health care and who doesn't. But since the government will no longer be allowing that decision to be made based on one's ability to pay, some other method of allocation will have to be found. Rationing might be the chosen solution, but I think probably, like Canada, we will end up with a system in which health care is allocated on the basis of waiting lists, political and personal connections, and the black market.[8] In other words, there will still be people getting left out, but now they will be the people who have to wait too long on the waiting list because they don't have the political connections to jump to the head of the line (Congressmen, of course, will get instant service) and because they can't use the black market, being either too law-abiding, too poor, or simply too ignorant of where black market medical care can be found. Now, if the national health care enthusiasts are explicitly saying that they prefer the Canadian method of allocation (plus a black market) to the American, then they aren't being irrational.[9] But if, as seems to be the case, they are under the impression that nationalizing health care will "assure adequate and affordable health care for all Texas citizens," then they are in for a rude awakening -- for in that case, they will have confused surface causes with root causes.

iii. Formal factors

The formal factor results from the essential nature of something. Its primary areas of service are in art and in tuning the precision of our language. The second is much more important for clear thinking, but we'll discuss the first briefly, since it helps illustrate what I mean by formal cause.

If you ask me, "Why are you giving that table one leg in the middle instead of a leg at each corner?" I can answer, "Because I like tables that are built that way." But what do you say to a person who says, "Why are you making that three-legged stool have three legs?" The obvious answer is, "Because if it had four legs, it wouldn't be a three-legged stool!" Why do three-legged stools have three legs? Because that's what it means to be a three-legged stool. The stool insists on having three legs, or else it is ruined -- or else it's something besides a three-legged stool (an piano stool, perhaps). And when we appeal to the essential nature of something as a reason for its being the way it is, we appeal to a formal factor.

Turning to art: if you ask, "Why did Cézanne use that shade of cool green in his landscape instead of a hot fluorescent pink?" the only satisfactory answer is, "Because hot pink would have ruined the painting." In other words, the painting itself demanded that shade of green. A good artist finds himself enslaved by his own creation; it takes over and starts acting like it has a life of his own. When that happens, the artist has created something that has its own form, its own nature, and if the artist tries to change it, he ruins the work of art. In a perfect work of art, every detail seems inevitable; we feel that it is somehow right. And when we can say, "Lord Peter couldn't possibly be converted at a Billy Graham crusade; he just isn't that sort of person," we are appealing to a formal cause. Why do you not cover a Chippendale chair with a vinyl seat cover? Because you'd be ruining the Chippendale chair. Why don't you put Sears siding on a vintage Victorian house? Because you'd be ruining a vintage Victorian house.

[N.B.--Do not include the rest of this discussion on art in the course; too much of a digression, with too few people interested in it, and besides there's even more of my personal opinion mixed in with it than usual. Of course artists frequently ruin their works of art. But that simply means that either they aren't very good artists, or else they were motivated by something else entirely -- maybe they wanted to make a political statement, or something. The very fact that we want to explain why the artist "messed it up" means that we have drawn a conclusion about the form the work ought to have had, and the thing that bothers us can't be reconciled to that form -- i.e., it cannot be explained by formal causes. Either you have an incompetent artist who couldn't see the form of his own work, or else you have an artist who has decided to stick some political commentary or something into his art, and thus is not, strictly speaking, functioning as an artist anymore.

This may not seem very important (who cares about art except critics?), and indeed you needn't spend hours contemplating theories of aesthetics in order to make good decisions about ordinary daily life. But it is surprising to see how frequently we see either that art is being treated as philosophy ("Renoir's paintings express the artist's conviction that beauty is the only real God") or else political propaganda is masquerading as art (e.g., the work of Robert Mapplethorpe). Of course, there is nothing to keep a work of art from being social commentary at the same time; The Grapes of Wrath is certainly far from "pure art," and yet it is unquestionably a great novel, and Sarte's Huis Clos is both important 20th-century drama and influential 20-century existentialist philosophy. But to talk about "what the author was trying to say" about politics is to discuss motivational or volitional causes, i.e., to treat it as philosophy or propaganda, while if we want to talk about whether or not it's a good play or a good novel, we have to treat its formal aspects. The Communists used to err by judging all art on whether it was "proletarian" (even if the artist had no political motivations whatsoever) while the far American left errs by insisting that "art" must not be criticized on the basis of morality (even if the artist is making a clear and offensive political or social statement such as that unquestionably made by "Piss Christ").

A last example: the whole plot of the movie Little Shop of Horrors, right up to the end, fits together and demands a tragic end. But at the movie's climax, just as the plant is about to destroy the foolish man who has underestimated the power of evil and tried to use evil for his own means, his house begins to collapse due to the violence of the evil alien plant, and out of the ceiling there falls a high-voltage line, apparently carrying several thousand volts. It is not explained why this high-voltage wire is running through the ceiling of a building. The wire is broken, but still live, and, conveniently, still complete with plenty of insulation for the hero to grab hold of. It obligingly falls just within his reach, but without hitting him or grounding itself through any of the rubble surrounding him. He snatches it up, grounds it on the plant, kills the plant, and then takes the girl and wanders off to their dream house in the suburbs, without any explanation of how they acquired it, or, if they already had it, why they hadn't left before. You cannot find a better example of the clumsy sort of ending known since Aristotle as the "god in a basket," the deus ex machina. Is this the result of incompetent playwrighting? Not at all -- in the stage version, if I am not mistaken, the plant destroys the protagonist, just as the plot calls for. So why did they change the ending in the movie? Isn't it because movie audiences like happy endings? -- i.e., because the producers were making entertainment rather than art and were willing to sacrifice artistic integrity for the sake of keeping audiences happy? That is, did they not allow business considerations (i.e., motivational concerns) to override artistic integrity (i.e., formal concerns)?]

But the most important application of formal causes is in the precision of our language. In talking about communication, I pointed out that we subconsciously organize our experience into bundles so that we can talk about it and think about it, and that different people and different societies organize the world in different ways. Now, the way we organize experience is shaped by the forms into which we fit our experience. If something can be fit into the form we call a tree, then we call it a tree. If something matches what we think of when we use the word murder, then we call it murder. But although our forms are somewhat arbitrary, they are not "created equal;" some are objectively better than others. And the precision and usefulness of a logical term is determined by how well it lends itself to formal causality.

Whew! Better use an example. I pointed out that you can't say that abortion ought to be illegal just because it's murder. The reason is that some people think illegality is associated with murder just as a general rule; it just so happens that murder usually is the sort of thing you ought to outlaw. But if we can show that the very essence of murder is that it can't be tolerated -- if we can show that the form of murder involves intolerable injustice -- then we can say of any murder, including abortion, that it ought to be illegal, on formal grounds. Then it will be completely responsible to respond to the question, "Why should infanticide be illegal?" with the answer, "Because it's murder." You will be able to cite a formal cause, because intolerableness is part of the very essence of murder, and if something can be tolerated, then it ipso facto is not murder.

But questions of form are questions of definition. For the meaning of a word is precisely the form to which it refers. So if we want to discuss the question of abortion, given the possibility of its being murder, we ought to take the following steps, in the order given:

1. We must first turn our attention to the essential nature (the form) of tolerableness. That is to say, what makes something tolerable rather than intolerable? Or, to rephrase the question, what makes something be the sort of thing that ought to be legal rather than illegal? Before we can do anything else, we must settle this question. (The reason abortion is such a controversy is simple: there is mass confusion about what sorts of things the government can and cannot outlaw.) We must define tolerable and intolerable with accuracy and precision.

2. Since "murder" is an emotional term never used in ordinary conversation (i.e., outside of the abortion controversy) except to refer to intolerable manslaughter, we should try to define the term to mean precisely that. That is, we should try to figure out what exactly it is that makes a killing intolerable, and then define murder so that it covers as many intolerable killings as possible without ever crossing over the line to cover an example of a killing that could be tolerated. In other words, we must trans-form our use of the term murder so that intolerableness becomes part of its essential form. And we must not confuse matters by using it in allegorical or imprecise ways. Once we have done this, we can reply to any question of, "Why should this or that murder be illegal?" by appeal to a formal cause: "Because it's murder; if it was the sort of thing that our legal system could tolerate, it wouldn't be murder." And the nice thing about formal causes is that they aren't the result of generalizations,[10] and therefore you don't have to worry about overgeneralization.

3. When we agree on the definition of murder, we turn to abortion, and here we must nail down what the form of abortion is. Is it abortion, for example, if a female fetus has to be removed from the womb in order to stop a life-threatening hemorrhage that neither she nor her mother could survive? If not, why? (Please note that both murder and abortion are ethico-moral terms, and therefore imply the presence of certain motivational and volitional factors -- as we will discuss shortly.)

4. When we agree on the definition of abortion, we then ask whether all abortions are murder by their very nature, i.e., by virtue of formal cause. If they are, then since all murders ought to be illegal -- and this due to formal causes, and therefore not vulnerable to overgeneralization -- and since all abortions are formally murder, then all abortions must be illegal. (Remember that we may have had to rule that some things we are accustomed to call murders or abortions are not essentially murder or abortion; in defining the terms precisely we may have had to restrict their use.) The legal question will be simply whether an abortion has taken place. But if abortions are not essentially murder, we go to steps 5 and 6.

5. If some abortions are not murder, we then must decide which are and which aren't, so that we may outlaw all those that are. But in such a case, the abortion will be outlawed not because it is abortion (that formal cause would be insufficient), but because this particular abortion happens to be murder (and that formal cause is sufficient).

6. Last of all, we must look at those abortions that are not murder, and consider whether there are any other reasons for or against outlawing them. But since any such reason will not be a formal reason, we will have to take these remaining abortions on a case-by-case basis in order to avoid the fallacy of overgeneralization.

If we compare this to the way the abortion issue actually is approached, we have two common approaches. The anti-abortion side says, "Stop murdering babies!" without having explicitly addressed the question of whether all forms of murder, and specifically those murders in which the victims have yet to be born, are intolerable. The pro-abortion side says, "Freedom of choice!" i.e., "Tolerate it! Tolerate it!," without having addressed the question of whether abortion is or is not tolerable. Any sober, responsible discussion will have to begin with the nature of tolerable-ness, and move from there through murder to abortion. I do not see such a discussion taking place.

My point in using this example is to show that formal causality can be a powerful tool of thought ‑‑ but only if words are clearly defined and used with precision.

iv. Motivational factors

Motivational factors bring us into the field of morality and ethics. Every moral or ethical statement makes assumptions about motivation. I once had a professor at Princeton challenge my class to come up with the most explicit physical definition of male homosexual intercourse that we possibly could, with the idea of defining the sin without recourse to motivation. Since this was a class full of Princeton college students, few of whom were particularly devout, conventional inhibitions were not on display, and we came up with a description that was explicit enough to make Dr. Ruth blush. Whereupon the professor managed to describe a scenario in which, we had to admit, all our conditions were filled and yet we couldn't accuse the participants of homosexuality. (Take my word for it; you don't even want to hear our definition, much less his scenario, but he did it.) That convinced me.

For a slightly tamer example, take adultery. Would you say that a married woman who of her own free will sleeps with a man who isn't her husband, commits adultery? Sure we would, right? (No doubt you realize this is a trick question and are refusing to go along with it.) Well, anyone who has seen the movie Overboard knows what the trick is: what if the woman has amnesia, and the man she's sleeping with has lied to her and convinced her that he is her husband? Is she committing adultery? You may ask, "Well, okay, but let's be serious here -- what are the odds of something like that really happening?" And your point would be valid, except that all it would prove is that, generally speaking, married women who sleep with other men are perfectly well aware that they are sleeping around on their husband, and so generally speaking women who do that are adulterers.

What we actually have in this case is a material definition of adultery -- that is to say, a definition that leaves out the question of motivational factors. Objectively speaking, she is deliberately sleeping with someone who, as it turns out, is not in fact her husband. But she thinks she's actually sleeping with her husband. In other words, while the material definition of adultery would be "deliberately sleeping with someone who isn't your spouse," an adulterer -- i.e., someone who was morally guilty of adultery -- would be someone who is "deliberately sleeping with someone who isn't his spouse," i.e., doing it and knowing that that's what he is doing. The motivation matters. (Christians will find a source of confirmation for what I'm saying in the fact that Jesus was so far from identifying adultery with its material characteristics that he could say, "If a man looks lustfully at a woman, he has already committed adultery with her in his heart.")

This distinction between a moral act and the ways in which that act is carried out is a source of confusion (at least, I hope it's confusion rather than cynical and dishonest manipulation of the facts) in the pro-abortion complaint that "those religious bigots think you women who have abortion are murderers." No doubt there are some who think so, the kind of anti-abortionist who stands around screaming, "You're possessed by Satan! You're going to burn in hell!" and makes the CBS Evening News as an example of the "typical" pro-lifer. But the basic legal issue is not whether abortion-choosing women are morally guilty of murder; that would presuppose knowledge about their motivation that we simply don't have. The question is whether abortion meets the material definition of murder (given, of course, proper definitions as discussed under formal causes). Abortion may, materially speaking, be murder (and therefore intolerable) without the woman and doctor in question having chosen to commit murder (in the moral sense). If they are honestly convinced that the fetus is not a human being, then they are at most guilty of negligent manslaughter, not murder. The propaganda of some pro-abortionists, and the intolerance of some anti-abortionists, blurs this crucial distinction.

That, by the way, is the other explanation of the flaw in the argument, "You pro-lifers don't really think abortion is murder, because you don't want to punish women who have abortions the way you would punish women who hired hit men to kill their two-year-olds." We used this argument as an example of overgeneralization,[11] but it also clearly depends on the confusion between material murder (that which is prohibited) and murder in the sense of moral responsibility (moral responsibility being an essential consideration in questions of punishment). After all, there aren't any million-dollar national organizations running around trying to convince impressionable young women that their two-year-olds are merely potentially human tissue masses. It takes a hefty helping either of depravity or of insanity to "abort a childhood," as it were; on the other hand, all you really need to abort a pregnancy is a normal American-sized helping of credulity. In short, you can call abortion "murder" (in the material sense), and therefore outlaw it, without calling 16-year-old incest victims whom Planned Parenthood talks into having an abortion "murderers," and therefore executing them.

Please note, however, that in the question of defining a moral act, you have to start with the nearest motivational cause. In Overboard, Annie/Joanna's motive for doing what she does with her body is "to make love to my husband." So she can't be called an immoral adulterer (since her confusion is not due to negligence on her part). But let's say that Joe Smith and his neighbor's wife Sarah Sugarbaker have "fallen in love," and Sarah has sworn to commit suicide if Joe won't run away with her and consummate the relationship. And let's say that Joe doesn't want to break up her marriage, but he gives in and runs away with Sarah "in order to save her life." Maybe he even hopes to use the episode to convince her to renew her commitment to her husband. Now, clearly there's nothing wrong with wanting to save another person's life, and it is extremely admirable to dedicate yourself to saving the marriage of the person you badly want to be married to yourself. But can he say that he is not really committing adultery, since his motivation is not "to commit adultery?" No, because "saving Sarah's life and marriage" is his motivation for committing adultery, but "making love to Sarah" is his motivation for sleeping with her. The nearest motivation is mutual sexual gratification, with someone he knows isn't his wife. So while we may argue about whether adultery is morally wrong in this case (a question of ethics), it's clear that, right or wrong, adultery is what he's committing. For the nearest motivation satisfies the definition of adultery.

A final point: the sense of the maxim, "Judge not, lest ye be judged," springs partly from the distinction between the material characteristics of a sin and the motivation behind it. Sometimes motivations are clear (at least the nearest motivations), as in the case of Joe and Sarah. But other times they are not. Certainly we can say, "If something is done with such-and-such a motivation, it is wrong;" that is why we can say that adultery or homosexuality or racial intolerance is wrong, since the definition includes the motivation. But it is another thing entirely to say that another person is actually driven by such-and-such a motivation (especially in view of the fact that we are so frequently deceived about our own). And I think that generally we presume far more than we are rationally justified in presuming about people's motives. However, it could just be that I personally have an unusually severe problem with judgmentalism, in which case I withdraw the "we" in the previous sentence.

v. Volitional factors

And now to volitional factors. Here we come to the question of responsibility. When we assign responsibility for something to someone, either in order to blame them or to praise them, we essentially are saying that what happened, happened because of some choice they made. In other words, they are a direct volitional cause of what happened.

It's a little more complicated than it may look, because you can be responsible for things you didn't choose to do; for example, a drunk driver is responsible for the death of any pedestrian he may run down, even if he "didn't mean to kill anybody." The fact remains that he made the choice to drive while drunk, and that choice resulted in a death, for which he is responsible, albeit by negligence rather than by malice.

There are two ways in which you can be a volitional cause of something. You can be an intentional volitional cause. That is to say, you effectively will an event -- you want it to happen, you try to make it happen, and you succeed. The second is by being an unintentional volitional cause. That is to say, you make a choice, and one of its consequences is the event in question, but that's not what you're trying to do.

When you assign responsibility for an event to someone, you assume that they were in some sense a volitional cause. If they were an unintentional volitional cause, you either assume that they knew the consequences of their action and chose anyway, or else that they didn't know but should have -- i.e., they were negligent. If they didn't know the consequences and couldn't have been expected to, we don't hold them responsible (even though they could still be considered, in a sense, an unintentional volitional cause).

Let's look at some examples, starting with intentional causes.

(1) I decide to shoot Jesse Jackson because I find him annoying. I am now a murderer, directly, maliciously and intentionally responsible for his death.

(2) A Neo-Nazi group hires me to assassinate Jesse Jackson, which I do. Both the Nazis and I are intentionally responsible.

(3) Neo-Nazi terrorists take over the Department of Defense and threaten the world with nuclear annihilation unless I drag Jesse Jackson onto the White House lawn and shoot him. The Rev. Jackson agrees to sacrifice himself for the good of the world. I pull the trigger and shoot him, weeping copiously all the while. I have allowed myself to be used as a material cause in the killing of Jesse Jackson. Since I could have refused to shoot him and allowed the terrorists to push the button, I am responsible for my choice; however, under the circumstances we probably would not say that I was guilty of murder. I did, however, choose to kill him, because that's what the demand was. (You can't say I didn't mean to kill him.) Note that the distinction between (3) and (2) is purely moral, i.e., it concerns my motivation; in both cases I am an intentional volitional cause. Example (3) is just unusual because I will his death without desiring it.

Examples of unintentional volitional causes:

(1) I decide to amuse myself by shooting bullets into the air in the middle of Washington, D.C. One of the bullets lands on Jesse Jackson and kills him. I am now guilty of manslaughter, directly and negligently (though unintentionally) responsible for his death.

(2) The Neo-Nazis of example (3) above demand merely that I shove Jesse Jackson off the top of the Washington Memorial. I do; as a result he dies. But I am an unintentional cause of his death, for I only intended to shove him off the Memorial; I was hoping all along that by some miracle he would survive (forlorn though that hope may have been). His death is an intended consequence for the Neo-Nazis, but it is an unintended (though foreseen) consequence for me. I am directly responsible for shoving him off (I willed it), but I am only unintentionally responsible for his death. Note that once you get into unintentional responsibility, you get into questions of negligence: in example (1) I am being negligent; in example (2) I am not being negligent. (I know, by the way, that this example seems absurd. It is deliberately so precisely because I’m trying to exaggerate the distinctions I want you to grasp, not because anyone would ever actually face such a choice.)

(3) I try to drive home drunk; I accidentally kill a small child. I am an unintentional cause; I am criminally negligent. I am the main volitional cause.

(4) I live in a small town, where it is well know that on Saturday nights at 2:00 a.m. I drive the three blocks between my house and the local bar, at twenty miles an hour, using a back road that nobody else uses at that time of night, after having had just enough alcohol to slow up my reflexes without destroying my ability to keep the car safely in my lane. A suicidal teenager lies in wait and leaps in front of my car; in my slightly inebriated state I am unable to stop in time to avoid killing her. Okay, I am a negligent unintentional volitional cause. But please note that the cause -- that is, the main cause -- is the teenager, whose decision came after mine (thus she was a nearer cause) and who directly willed the event (so her choice was intentional while mine was not). My negligence made it possible for her to use me to kill herself. Note how this alters the case from example (4), in which there was no intentional volitional cause.

(5) A high school girl goes to a college frat party wearing a miniskirt, a T-shirt that says, "This body is mine, but I share," and no bra. She is accompanied by a friend, who after a little while decides to leave; our heroine stays on alone. She has quite a bit more to drink than she can handle; perhaps this is why she enters the wet T-shirt contest, which she wins. One of the frat brothers offers to take her up to his room and give her a dry T-shirt. She goes up to his room and promptly passes out on the bed. Whereupon he rapes her. Now, is it her fault she got raped? Did she "get herself raped"? Is she, in short, directly responsible for her rape? Well, in a sense; her choices unintentionally put her in a situation where she could be raped. But here are some other pertinent questions. (a) Did she will her own rape? Absolutely not. So she isn't a intentional cause. (b) Did anybody else will her rape? Sure. Her rapist did. So the main blame is unquestionably focused on the rapist. Still, she's going to feel some guilt, so let's ask some more questions to see whether it's justified. Was she negligent? Yep. She was a little fool, in fact, and not a particularly moral one. Does that make her deserve to be raped? Nope. The "punishment" is all out of proportion to the event. Furthermore, is she the nearest volitional cause? No, that would be the rapist. Was she a necessary cause? Well, yes; if she'd had sense, he wouldn't have had the chance to rape her. Was she a sufficient cause? Not at all. She trusted him to treat her responsibly, and a woman ought to be able to trust men to treat her responsibly; if her rapist had acted responsibly, she'd not have been raped. She can't in fact trust strangers that much, but she ought to be able to.

What her guilt amounts to, in fact, is this: she negligently, but unintentionally, put herself in a bad situation. That situation then became the (admittedly necessary) situational cause that made it possible for her rapist to choose to rape her. Should she feel guilty? Sure, to a degree ‑‑ that is, she should regret being stupid and take pains not to be stupid in the future. Should she feel guilty because she got raped? Nope. Should she feel like she deserved it? Nope. Should she feel like she "got herself raped"? Nope. The extent of her responsibility is that she foolishly got herself into a vulnerable situation, a consequence of which was that a rapist raped her. She was an unintentional volitional cause of the situational cause of her rape. That is a very different thing from being a self-rapist, as it were.

Therefore we must conclude that while she is responsible for acting foolishly, she is not responsible for being raped. The responsibility for her rape rests entirely on her rapist. He, and he alone, qualifies as a direct volitional cause, and therefore he, and he alone, is guilty of her rape. No doubt she will feel guilty and responsible for her rape as well as her folly, but that exaggerated guilt will be the result of the emotional damage he has caused her, not a reflection of any real responsibility for the rape. And he may well try to say that she "asked for it." But that will simply be a vain attempt to pass off his responsibility for his actions.

(The same considerations apply to the Princeton University Women's Center's claim that, "When you laugh at a sexist joke, you give every man present permission to commit rape." You don't have to have been through the previous discussion to know that this statement is absurd, but the discussion helps explain why it is absurd.)

To sum up: there are intentional volitional causes and unintentional causes. Intentional volitional causes (people who deliberately choose to do something) are always responsible for their actions. Unintentional volitional causes (people who do something without meaning to) are responsible for the consequences insofar as they foresaw them or were morally obliged to foresee them (and therefore were negligent.) The question of where you pass from the "acts of God" class (like a meteor landing on a car after the wife insisted, over the husband’s objections, to stopping and asking for directions) to the "criminal negligence" class (the drunk driver killing a toddler) is not an issue for this class, thank heavens.

And that, at long last, concludes our exploration of the wide variety of meanings that are carried by the single, ambiguous word cause. Just as a final example of the way disastrous conclusions can be reached purely through the confusion of one kind of cause with another, I now quote a passage from Dorothy Sayers's essay on education called, "The Lost Tools of Learning."

Ms. Sayers is expressing her alarm at the many dismaying consequences of an educational system that does not teach its pupils either how to think or how to learn, and she is listing some examples of said dismaying consequences.

...It is more alarming when we find a well-known biologist writing in a weekly paper to the effect that: "It is an argument against the existence of a Creator" (I think he put it more strongly, but since I have, most unfortunately, mislaid the reference, I will put his claim at its lowest) -- "an argument against the existence of a Creator that the same kind of variations which are produced by natural selection can be produced at will by stock-breeders." One might feel tempted to say that it is rather an argument for the existence of a Creator. Actually, of course, it is neither: all it proves is that the same material causes (re-combination of the chromosomes by cross-breeding and so forth) are sufficient to account for all observed variations -- just as the various combinations of the same 13 semitones are materially sufficient to account for Beethoven's Moonlight Sonata and the noise the cat makes by walking on the keys. But the cat's performance neither proves nor disproves the existence of Beethoven; and all that is proved by the biologist's argument is that he was unable to distinguish between a material and a final [i.e. motivational] cause.[12]

We now turn to the fallacy of false cause properly so called

2. False cause (strict sense)

In almost every case, the fallacy of false cause results either (a) from the available information's being inadequate for detection of a hidden cause, or (b) from taking two things that are related in some way and deciding that they are related as cause and effect, when in fact they are related by togetherness, temporal succession (one comes after the other), statistical correlation, reversed causality (we've got them backwards), double effect of an unidentified cause, or mutual influence. To avoid (a), you just need as much information as possible. To avoid (b), the general principle is simply this: you have to be able to explain how the cause produces the effect. If you can't explain how it is that F.D.R. caused the Great Depression to end, then you don't have the authority to say, "F.D.R. brought us out of the Great Depression."

Americans generally do not insist on understanding how one thing causes another before accepting someone's assertion that the former in fact does cause the latter. That fact in itself is sufficient to explain a great deal about American politics. It explains (if I may allow personal opinion to color my examples for a moment) the general public's acceptance of the bromides of Keynesian economics, despite the fact that Keynesianism is both a theoretical and empirical failure. (And the practical effect on the American economy may be easily guessed at when one considers the fact that both major political parties are Keynesian to the bone.) It explains the fact that it is possible to find college students who can give no better reason for their voting decisions than, "Well, the Republicans have been running things for twelve years, and I don't like the way the economy is now, so I'm going to give Clinton a try," or, "Well, the last time we had a Democratic president, it was a disaster, so I'm not going to trust Clinton." And then there is the national health care debate, which is one huge glaring example of a false cause fallacy.

Not that it's just American politics; all 500 pages of Das Kapital may be summed up in the words, "Give economic 'classes' false reality, posit a 'class war,' and falsely assert that the class war is the cause of all social problems." My point is that any well-disciplined mind refuses to assert cause and effect unless it can give at least some explanation of how the cause produces the effect, and that such discipline does not come easily to a person raised in our culture.

Let us look in detail at each of the kinds of relationships that can be mistaken for cause and effect.

a. Hidden causes

This occurs when we lack the necessary information to figure out the cause. The gullible country boy thinks the magician is psychic; in fact the magician's brother came into town a week ago and did a bunch of research. The cuckold thinks his wife isn't answering the phone because she's gone shopping; in fact she isn't answering the phone because she's entertaining his best friend. A man thinks his friend the stock market trader is a financial genius; in fact his friend is enthusiastically indulging in illegal insider trading.

The only way this can work is if a plausible candidate for the cause exists. A person who doesn't believe in ESP will assume that the magician is cheating somehow; only a man who knows that his wife sometimes goes shopping will explain her absence as shopping; and a man who knows that his friend failed economics twice in college will figure something funny's going on if the guy starts cleaning up in the stock market.

So how does a false cause become a plausible candidate? We'll now look at the most common ways.

b. Confusion of correlation and cause

In my list of kinds of relationships, the first three were togetherness, temporal succession, and statistical correlation. All three of these are actually variants of the general concept of correlation, but it's worth examining each separately.

Correlation basically just means that we see two things that we happen to associate with each other. They are correlated; they go together. Unfortunately the human mind has a remarkable propensity to leap from the perception of correlation to the assumption of causality, thus committing the fallacy of false cause.

i. Togetherness (propinquity): "With it, therefore because of it"

"These two things are together; therefore the first caused the second." That is the reasoning involved in this form of the false cause fallacy. For example: "A Republican was president when we entered the Great Depression. A Democrat was President when we came out. Therefore the Republicans caused the Depression, and the Democrats got us out of it." We would think someone was a little biased if they blamed the referees for their wins and losses in a basketball game -- "when we had Smith and Jones, we won by 15 points, but when we got those worthless Cook and Tanner, we lost by 15" -- and we think baseball coaches who wear the same socks ten games in a row because "every time I wear these socks, we win" are being silly as well as repulsive (unless they wash them in between games, I guess, in which case they're just silly). But we appeal to such arguments all the time, when it suits our prejudices to do so.

Actually, such an argument is so silly that we have to have a predisposition to the view we're trying to establish. Obviously the person in the first example blames the Republicans for the Depression, in all probability, because he's a Democrat; if he were a Republican, he would probably assign the cause of the Depression to some vague factor such as immorality and overspeculation in the Roaring Twenties, and claim that the cause of the recovery was World War II. (In fact, a reliable way to tell whether a person is a Republican or a Democrat is to ask them what or who got us out of the Depression, and to see whether they say, "World War II," or, "President Roosevelt.") Furthermore, the first person is taking advantage of the fact that Americans have the habit of blaming Presidents for the state of the economy, without asking themselves how exactly the President affects the economy or what role Congress might play, while the second person is taking advantage of the fact that Americans have vague recollections of being told by some trustworthy authority figure (probably a high school civics teacher) that holding a war helps the economy (if you win). But ask the Democrat for precise details on the economic effect of F.D.R's policies, and you will be fortunate if he can even tell you what those policies were, much less how they went about affecting the economy. And ask a Republican how a war helps the economy, and you will be lucky to get anything more precise than the idea that war makes the government spend a lot of money and stimulates business, if you get that much. If you ask him to explain what makes massive war spending good for the economy but massive social spending bad for it (since he presumably doesn't support the Democratic ideal of the welfare state), the odds are that you will not get a very valuable response.

Let me emphasize that the problem is not that the Democrat credits F.D.R. or that the Republican credits World War II. The problem is that of the Democrats who say F.D.R. caused the recovery, only a small fraction can give an intelligent explanation of how he caused the recovery; and that of the Republicans who say World War II caused the recovery, only a small fraction can give an intelligent explanation of how that causation worked. And such people are committing a form of the false cause fallacy: they are confusing togetherness with cause and effect.

The power of the togetherness fallacy comes, I think, from two facts: first of all, we are not comfortable in a random world, so we look for cause and effect simply because it's psychologically comforting to be able to attribute a given effect to a particular cause; secondly, we want certain causes to have particular effects, and when we see the desired "effect" together with the desired "cause," it is extremely tempting to assert causality. That's why it is primarily Democrats who try to convince themselves that Roosevelt's policies ended the Depression, and primarily Republicans who try to convince themselves that it was really World War II, not Roosevelt. But what if we have these two factors predisposing us to attribute some particular effect to a particular cause, and it turns out that not only are they close to each other, but the "effect" comes right after the "cause"? After all, we expect effects to follow causes, and here's the effect, following the cause just like it's supposed to. What you get, in that case, is an even more potent form of this confusion, a form known as the post hoc fallacy, to which we now turn.

ii. Post hoc (temporal succession): "After it, therefore because of it" (post hoc ergo propter hoc, or post hoc for short)

The post hoc fallacy (for this one, I'm going to use the short Latin tag, since the best English I could come up with was "temporal succession"), like most, is easy to understand and seems too stupid to convince anyone. Yet it gets committed all the time. The fallacy runs like this: "Every day, my rooster crows, and then the sun comes up. So my rooster's crowing must be what makes the sun come up." Pretty silly, huh? But how about this: "I prayed that God would heal my cold, and it got better. It's a miracle!" The logic is the same as that in the rooster-sun example; it is not at all convincing to anyone who doesn't already firmly believe that there is a God who goes around healing people's colds. (In fact, Christians who give "testimonies" of this sort in order to convert atheists frequently discover that the atheist's reaction is amusement or disgust ‑‑ since one of the properties of the post hoc is that it's easy to spot when other people do it, as long as you don't share the prejudice that has set them up for it.) Post hoc is the fallacy that supports every superstition known to man: "A black cat crossed my path one time, and the very next day my wife left me." "I had a no-hitter with two out in the ninth, and then the bat boy said something about it, and the next pitch the batter homered." Despite the fact that it is so clearly silly when applied to the rooster and the sun, it is a practically ubiquitous fallacy.

My favorite example comes from Lyndon LaRouche's propaganda rag The New Federalist. LaRouche is violently opposed to the North American Free Trade Agreement, which he seems, if I understand him correctly, to consider a genocidal plot to wipe out the Mexican race. Several months ago the Federalist ran a front-page article headlined "Auschwitz on the Rio Grande," the "Auschwitz" in question being the maquiladoras built by American companies just across the border in Mexico. We were assured by the reporter that NAFTA will lower Mexican wages, not raise them. Now this defies the law of supply and demand, which is valid if anything in economics is valid. So it's a pretty startling thing to maintain. What was his proof? Well, when the maquiladoras were being built, people said Mexican wages would go up. But Mexican wages are much lower now than they were before the maquiladoras were set up. Therefore the maquiladoras made wages fall, through their evil habit of making more jobs available to the Mexican laborers! The minor points of the intervening devaluation of the peso and collapse of the oil industry are ignored -- the maquiladoras were built, and then wages fell, so the maquiladoras clearly make wages fall.

Of course, the post hoc fallacy doesn't get committed just because you say one thing happened after another and was therefore caused by it. If the Christian evangelist were to cite a cure which cannot be explained by material, nonmiraculous causes, then you would have an entirely different matter. (Though even then, a thorough-going materialist will prefer to say that the matter is "inexplicable" rather than say that a miracle is an acceptable explanation.) The fallacy consists not in asserting that one event was caused by an earlier one, but by asserting it with no grounds other than the fact that the "cause" was earlier than the effect.

iii. Statistical correlation

We began this section with simple togetherness, which is bad grounds for asserting causality. We moved on to togetherness-plus-temporal-succession, which is better than simple togetherness, but not much. We now strengthen the cause-effect case slightly more by entering the realm of statistical correlation.

I have no intention of teaching a course on statistics. That is a specialized branch of study that is simply not within the scope of this course. I merely want to say four things about statistical correlation, in order to further our discussion of causality.

First, statistical correlation is just seeing two things together over and over and over again. If Ann Richard's granddaughter wins the lottery, we have a coincidence. If she wins the lottery five times in a row, or if all of Gov. Richard's granddaughters win the lottery consecutively, we begin to suspect that some untoward form of causality is at work, for we are perceiving a statistical correlation between winning the lottery and being Ann Richard's granddaughter.

Second, there is a difference between significant correlation and insignificant correlation, and the difference is simply that the more times you see two things together (in statistical terms, the larger the sample size), the less chance you have of being fooled by sheer coincidence (in statistical language, the greater the confidence level). If you flip a coin five times, and it comes up heads every time, you still have a fifty-fifty chance of it coming up tails, because there's nothing particularly strange in getting five heads in a row. But if you get five hundred heads in a row, you'd better not bet against what is obviously a loaded coin. What constitutes significant sample size in any situation is something you'll have to study in statistics; what matters here is just the general principle that the more examples you have, you more sure you can be that you don't have mere coincidence on your hands.

Third, statistics presume that the deck is not stacked. P. J. O'Rourke has pointed out the folly of those (including himself) who trusted opinion polls taken in El Salvador before the election that drove the Sandinistas out of power. In a country where certain opinions are illegal and can get you killed, results of opinion polls are likely to be biased toward legal opinions. Or take the famous case of the paper which announced that Dewey would win the election against Truman based on its polls of its readers; it failed to take into account the fact that the majority of its readers were wealthy businessmen and therefore skewed toward the pro-business candidate. Stacked decks of this sort invalidate statistics.

Fourth, a significant correlation merely shows that some form of cause and effect is probably at work. It does not prove that the "independent variable" is the cause and the "dependent variable" is the effect. Take, for example, the study that showed that frequency of rape in large cities varies directly with the consumption of ice cream. This hardly demonstrates that we should outlaw ice cream in order to stop rape, or that Baskin Robbins should promote rape in order to stimulate sales. There are several different ways in which correlation can be confused with cause and effect, and in a moment we will look at them: reversed causality, double effect, and confusion of cause and mutual influence.

As an aside, let me make one final note about statistics: the only things easier to lie with than statistics are "scientific studies," as Mark Twain pointed out when he said, "There are three kinds of lies: lies, damned lies, and statistics." Until you have (a) studied statistics or research methodology and (b) assured yourself of the method used to generate any particular set of statistics or scientific study, you should probably just distrust them on general principles. We'll have more to say about this when we get to selecting authorities in the third part of the course.

c. Reversed causality (confusion of cause with effect)

In reversed causality, you get the cause and effect backwards. For example, the sun doesn't come up because the rooster crows; instead, the rooster crows because the sun's about to come up. One of my great-uncles, a chronic failure, used to say, "I'll never succeed in business because I'm snake-bit; I've always had bad luck, so I know it'll happen again as soon as business picks up." In fact, of course, his "bad luck" was a self-fulfilling prophecy, a result of the fact that he expected to fail and reacted in ways that made his expectations be fulfilled. The members of my high school basketball team used to complain about how other teams' home referees were always causing us to lose games. But I've only been in one game where I believe that the referees actually caused us to lose. We lost a whole bunch of games because if the ref blew one call, half our players would decide that the referees were cheating us, would quit concentrating on basketball and start abusing the "cheating refs," get the officials mad at them in return, and wind up blowing the game -- because the team that gets the breaks, including favorable calls, is the team that's hustling and playing smart. We didn't lose the game because the calls weren't going our way; instead, the calls weren't going our way because we weren't playing hard enough and smart enough to get the breaks, being too busy assuring ourselves that the refs were trying to make us lose. We had the causality reversed.

d. Double effect

In double effect, one thing is seen as a cause of something else, when in actual fact both things are caused by a common root cause. Here's where the ice cream/rape link falls: when it's hot, ice cream is more attractive and opportunities for rape are increased, while in the middle of a blizzard you are unlikely to see rapists lurking in the alleyways or children eating ice cream cones in the streets. This example makes the fallacy so clear that I don't think we need any more examples.

e. Confusion of cause and mutual influence

We confuse cause and mutual influence when we conclude that one thing causes another when in fact they feed on each other, in what is frequently referred to as a "vicious cycle." Far and away the best example -- though I hate to use it, in a way, because of possible reactions from Black activists -- is racism in America. Any White person who has been to a good-sized college knows that the vocal part of the Black community spends a lot of its time talking about how evil White people are. A basic tenet of Black activists at Princeton while I was there was, "All Whites are racist." Now, either this is false, or else it is equally true that, "All Blacks are racists," but the second would have been vehemently denied by the same Black activists. Or you can take, as an example of the attitudes typical to this type of Black activist, the response of a Black New York talk show host when he was asked whether his show was driven by Black hatred of Whites: "Hate? No, hate is something White people do." I regret to say that this seems to be Jesse Jackson's attitude as well; I doubt that even his supporters would attempt to deny that he has a flagrant double standard when it comes to judging the behavior of Whites and Blacks (though they would maintain that the double standard was justifiable). So the average White person knows that the Blacks who make all the noise seem to have a pretty deep and abiding hatred of White people, or at the very least that they see the world as being the people of color vs. the people of no color. (This would not have been true when M.L.K. was alive, but we no longer have M.L.K.; we have Jesse Jackson instead.) Now what is the average White person's reaction to this? Simple: he begins to assume that any given Black person hates him until proven otherwise. In other words, he becomes a racist.

But is this to say that racism in America is all the Black activists' fault? Of course not; that would be absurd. The problem is that racism begets racism. White racism causes understandable Black bitterness; Black bitterness tempts Black people to racism; Black racism understandably repels and angers Whites who aren't yet racist; their anger and repulsion pushes them toward racism; and on and on the cycle goes. The question ceases to be whose fault it is or who the "real" racists are. The question becomes simply how to break the vicious cycle of bitterness and hatred.

Self-fulfilling attitudes are another example of this kind of loop. The person who expects White people to hate him behaves in ways that are pretty much guaranteed to make any normal person hate him; he acts differently toward Black people; the result is that White people hate him and Black people don't, which confirms and strengthens his expectation that White people will hate him. The person who thinks Black people are congenitally rude treats Black people rudely, with the result that they are usually rude in return, which emotionally confirms his conviction that Black people are congenitally rude. The person who is dominated by fear of failure is caused to fail by his fear, but then the failure confirms and strengthens his fear, leading him even more likely to fail the next time. Or it can work the other way: the person who expects to succeed, succeeds, which makes him expect it even more firmly the next time. The more you sin, the more sinful you become, and the more sinful you become, the more likely you are to sin. The more you exercise, the healthier you get, and the healthier you get, the more you enjoy exercising. You get the idea.

We have not exactly exhausted the topic of causality; if this were a course on decision-making or problem-solving, we would have to go into yet more depth. But fortunately this is not such a course, and we can consider this sufficient and move on to other things.

[1]The term science has two meanings, the modern and the pre-modern. In the pre-modern terminology, a "science" was any body of reliable knowledge. Thus theology was for Aquinas the most (objectively) reliable science, insofar as it was Divinely revealed, because Divine omniscience never errs. The "scientific" revolution, however, has taught us to consider no body of knowledge reliable unless derived by the "scientific method." The extreme form of this doctrine is found in positivism; more commonly one sees its results in symptoms such as the tendency to look upon physicists as reliable sources of theological opinion. I am perfectly willing to restrict my use of the term science to the modern denotation, i.e., to "hard" sciences that rely on experimental evidence rather than pure reasoning. But, as we shall see, this means (a) that some fields of study, such as economics and much of psychology, are not "sciences" in this sense; (b) that such "non-sciences" are nevertheless quite reliable if their methodology is objective and appropriate for the field of study in question; and (c) that much of the knowledge that is crucial to our success in life cannot be arrived at "scientifically." Thus for me, "non-scientific" is not a pejorative term; it does not at all mean "useless," "invalid," or "unobjective."

[2] In an amusing comedy routine by The Geezinslaws (a local Austin country-music-and-standup-comedy duo whose live albums I highly recommend), Sammy relates that he saw Son one day looking terribly depressed. "What's the matter, Son?"
"You know that girl I was in love with?"
"Sure."
"Well, I found out she was just a golddigger. She only wanted me for my money."
"But, Son...you don't have any money."
"Yeah, she was stupid, too."

[3]I urge any psychologist to remember that I use the term science in a very precise and limited sense, and that in implying that psychology is to some degree non-scientific I do not mean anything that be construed as an insult to the field. I can't afford to; my wife has a degree in psychology.

[4] Borrowed shamelessly from C. S. Lewis's Mere Christianity, I might add.

[5]Actually, this statement is not precisely true. But when psychology turns to dealing with choice, it must to some degree abandon the scientific method. For example, in a particle accelerator, the electrons are not aware of the fact that they are experimental subjects and do not change their behavior accordingly. But subjects of psychology experiments are often unavoidably aware of their role as subjects and may try to outguess the experimenter, thus altering their behavior. And even when the subjects do not realize that they are being observed, the most the experimenter can do is to arrive at statistical conclusions, e.g., "Most men, given a choice, will watch Monday Night Football instead of Murphy Brown." That does not mean that any particular man will choose football over comedy. Thus psychology is restricted to statistical observations and cannot generate laws such as, "The force of gravitational attraction between two bodies is inversely proportional to the square of the distance between said bodies." After all, the planet Jupiter cannot take it into its head to start attracting its moons with twice the normal force, or to repel Io because it is offended by Io's behavior.

All of which simply means that while the psychology of choice may be perfectly valid, it is not (in our sense) "scientific."

[6]For those familiar with Aristotle, I am merely saying that conditions cannot be efficient causes.

[7]The stated goal of one lobbying organization, as expressed by their representative at the door of my home in Austin.

[8]At the moment, Canada does not have a black market. Instead they have the American health care system, which functions as a sort of legal black market for Canadians: Canadians who opt out of the Canadian system simply drive across the border, making "back alley" medicine unnecessary. If the U.S. health care system becomes a copy of the Canadian, a U.S./Canadian black market will then develop, since there will be no even relatively free and sophisticated market available to anyone who isn't rich enough to fly to Europe but who wants to circumvent the problems of the legal, nationalized health care system.

[9]Except insofar as they fail to see that the U.S. cannot have the Canadian system, since there is no alternative market to perform for us the role that we currently perform for Canada, viz. the role of a safe and legal "black market."

[10]Instead they are the result of definition.

[11]Well, actually, we didn't, but I meant for us to, and it would have been a good example.

[12]From A Matter of Eternity: Selections from the Writings of Dorothy L. Sayers (Grand Rapids, MI: Eerdman's, 1973), p. 112.