ARE THERE RIVAL CAUSES?
We begin this chapter with a story.
An inquisitive little boy noticed that the sun would show up in the sky in the morning and disappear at night. Puzzled by where the sun went, the boy tried to watch the sunset really closely. However, he still could not figure out where the sun was going. Then, the boy also noticed that his babysitter showed up in the mornings and left at night. One day he asked his babysitter where she went at night. The babysitter responded, "I go home." Linking his babysitter's arrival and departure with the coming of day and night, he concluded that his babysitter's leaving, caused the sun to also go home.
This story clearly illustrates a common difficulty in the use of evidence: trying to figure out what caused something to happen. We cannot determine an intelligent approach to avoiding a problem or encouraging a particular positive outcome until we understand the causal pattern that gave rise to the phenomenon in the first place. For example, we want to know what caused the steady rise in oil prices in the United States over the last few years. Or, why the suicide rate among professionals increased over the last 10 years?
The story also shows a very common difficulty in using evidence to prove that something caused something else—the problem of rival causes. The fictional tilde boy offered one interpretation of his observations: the sun sets at night because my babysitter goes home. We expect that you can see another very plausible explanation for why the sun sets. Although rival causes will rarely be as obvious as they are in our story, you will frequently encounter experts presenting one hypothesis to explain events or research findings when other plausible hypotheses could also explain them. Usually, these experts will not reveal rival causes to you because they do not want to detract from the sound of certainty associated with their claims; you will have to produce them. Doing so can be especially helpful as you decide "how good is the evidence?" The existence of multiple, plausible rival causes for events reduces our confidence in the cause originally offered by the author.
Searching for rival causes will always be appropriate when a speaker or writer presents you with some evidence and offers a cause to explain it.
Critical Question: Are there rival causes?Attention: A rival cause is a plausible alternative explanation that can explain why a certain outcome occurred.
When to Look for Rival Causes
You need to look for rival causes when you have good reason to believe that the writer or speaker is using evidence to support a claim about the cause of something. The word cause means "to bring about, make happen, or affect." Communicators can indicate causal thinking to you in a number of ways. We have listed a few.
X has the effect of. . . X deters . . .
X leads to . . . X increases the likelihood . . .
X influences . . . X determines . . .
X is a factor in . . . X contributes to . . .
X is linked to . . . X is associated with . . .
These clues to causal thinking should help you recognize when a communicator is making a causal claim. Once you note such a claim, be alert to the possibility of rival causes.
The Pervasiveness of Rival Causes
On the afternoon of March 28, 1941, Virginia Woolf wrote two letters, sealed them, and placed them on the mantle. She quickly put on her coat, grabbed her walking stick, and headed outside. She crossed the meadows to the river Ouse, where she put large stones into her coat pocket and threw herself into the river, committing suicide.
Authors have offered numerous hypotheses to explain this event, including the following:
1. Virginia Woolf had a fear of impending madness. She had a history of mental illness and depression. Also, given the recent outbreak of World War II and her history of an inability to deal with aggression, Virginia decided it would be best if she took her own life.1
2. Some psychologists argue Virginia Woolf had an intense attachment with her father. The attachment was so strong that Virginia developed to be much like her father in many significant ways. One important event in Virginia's life, therefore, was watching her father's deteriorating health when Virginia was in her early twenties.
She never forgot about her father's suffering and deterioration. So, when Virginia was 59 years old and feeling that her writing was beginning to deteriorate, she took her life in order to avoid identifying with, and in essence becoming, her dying father.2
3. Virginia Woolf was disillusioned with her marriage to Leonard Woolf. The two had a sexless marriage, and Virginia found companionship in an extra-marital lesbian relationship. Virginia's homosexuality put tremendous strain on the marriage that was in turn only made worse by the fact that Leonard probably had several affairs. Virginia was a very jealous person and did not take lightly to these affairs. Her dissatisfaction with the condition of her marriage led her to take her life.3
Woolf s own writing leads to any number of possible causes for her suicide. Now, let's leave Virginia Woolf's suicide for a moment and examine a different event in need of explanation—the findings of a research study.
A researcher reported that eating celery helps curb aggression. 151 women were surveyed, and 95 percent who reported eating celery on a regular basis also reported low levels of aggression, or overall irritability. Of the portion of women who do not eat celery on a regular basis, 53 percent reported frequent feelings of irritability, agitation, and aggression.
In this study, the researcher probably began with the hypothesis that eating celery causes reduction of aggressive impulses, and he found evidence consistent with that hypothesis. But let us offer several rival, or different, causes for the same findings.
1. Research participants were highly suggestible, and the expectation of low levels of aggression was responsible for the reported differences; like the sugar pill placebo effect in medicine, believing that eating celery lowers aggression might have stimulated a number of physical and mental processes that caused participants to feel less aggression.
2. Participants wanted to please the researchers; thus, they reported feeling low levels of aggression, even though they did experience some aggressive feelings.
3. Nothing is known about the women involved in the study. It is entirely plausible that those who eat celery are health conscious, and thus are more likely to exercise. The increased amounts of exercise can be an outlet
for aggression, and thus lower feelings of aggression. Those who do not eat celery regularly may not exercise as often and thus do not have an outlet for their aggression.
Now, let's leave the research laboratory for a moment and move to the national pages of our newspapers and examine an argument related to crime statistics.
Since 1993, the levels of serious violent crime in the United States have decreased steadily. It is obvious that the heavy focus we place on law enforcement is no longer necessary. People are becoming civic minded and are choosing to no longer pursue a life of crime. Money spent on law enforcement can now better be spent elsewhere.
The hypothesis offered by the writer is that people's increasing civic engagement is the cause of the decrease in violent crimes over the last 12 years. But, let's again generate some plausible rival causes:
1. Violent crime rates have decreased because of the increased focus on law enforcement the writer is specifically calling to be cut. An increased concern with law enforcement, and not the civic concerns of citizens, caused violent crime levels to decrease.
2. Recent legislative actions have increased the punishments associated with violent crimes. These increased punishments make the costs of committing a violent crime far outweigh the benefits of committing violent
crimes. People are not more civic minded, rather they are looking out for their own personal interests.
3. The booming economy in the 1990s could have decreased the number of people in poverty. Given that the poor are typically the perpetrators of what we call violent crimes, fewer poor people would lower the violent crime rate.
Now, let's examine some important lessons that can be learned from Virginia Woolf s suicide, the celery research study, and the crime statistics.
In the following sections, we explore the implications of these lessons for the critical thinker.
Detecting Rival Causes
Locating rival causes is much like being a good detective. When you recognize situations in which rival causes are possible, you want to ask yourself questions like:
The Cause or A Cause
The youth are exhibiting an alarming increase in the rate of depression among elementary aged children. Talk show hosts begin to interview the experts about the cause. It is genetic. It is the prevalence of teasing among
peer groups. It is parental neglect. It is too much TV news coverage of terrorism and wars. It is lack of religion. It is stress. The experts may claim to have the answer, but they are not likely to know it. That is because a frequently made error is to look for a simple, single cause of an event when it is really the result of a combination of many contributory causes— a cause that helps to create a total set of conditions necessary for the event to occur.
Multiple contributory causes occur more often than do single causes in situations involving the characteristics or activities of humans. In many cases, the best causal explanation is one that combines a considerable number of causes that only together are sufficient to bring about the event. So, the best answer experts can give to the talk show hosts' question is "We don't know the cause for such events, but we can speculate about possible
causes that might have contributed to the event." Thus, when we are searching for rival causes, we need to remember that any single cause that we identify is much more likely to be a contributory cause than the cause.
When communicators fail to consider the complexity of causes, they commit the following reasoning fallacy:
F: Causal Oversimplification: Explaining an event by relying on causal factors that
are insufficient to account for the event or by overemphasizing the role of one or more of these factors.
In some sense, almost all causal explanations are oversimplifications; thus you want to be fair to communicators who offer explanations that do not include every possible cause of an event. Causal conclusions, however, should include sufficient causal factors to convince you that they are not too greatly
oversimplified, or the author should make clear to you that the causal factor she emphasizes in her conclusion is only one of a number of possible contributing causes—a cause, not the cause.
Rival Causes and Scientific Research
Scientific research attempts to isolate some of the most important contributing causes from other extraneous causes and provides a major source of hypotheses about what causes events in our world. Researchers start with tentative beliefs— hypotheses—about causes of events. For example, when a massive wave killed
thousands of people, researchers generated many hypotheses about the cause of tsunamis.One hypothesis was that tsunamis are caused by massive underwater earthquakes. Once a hypothesis has been firmly established by dependable research evidence, it changes from a hypothesis to a law. In the domain of complex human behavior, however, there are very few established general laws. Stated claims like "fundamentalism causes terrorism," and "tax cuts cause economic growth" sound like laws, but we need to remain skeptical of the generalizability of such claims. They must currently be viewed as hypotheses, not laws, and are best stated as follows: "fundamentalism may be a contributing cause in the decision to resort to terrorism," and "tax cuts may be a contributing cause in stimulating economic growth."
Then, what should you do when speakers or writers use findings from research studies to conclude that one event causes another? First, remember that their conclusion should be viewed as a cause, not the cause. Then try to find out as much as you can about the research procedures used to produce the findings that support the hypothesis. Finally, try to determine rival causes that might explain the findings. The more plausible rival causes that can account for the findings, the less faith you should have in the hypothesis favored by the communicator.
Let's use the following argument to practice detecting rival causes.
Playing violent video games for long periods of time appears to increase the likelihood that a child will physically assault another child. The results confirm the general suspicion that violent video games cause violence in children. The relevant research findings are from the Center for Preventing Youth Violence, which enrolled 1,001 male children from across the United States. One-third of the children played "violent" video games, one-third played "non-violent" video games, and the remaining third did not play any video games. The children played several hours of video games alone every day for two weeks. At the end of the two weeks, the children from the different groups were put into a room with toys so that they could play together. Those children who played violent video games were more likely to get into physical altercations with other children than were those who played "non-violent" video games, or no video games at all.
Should parents take away all of their children's "violent" video games?
Not until they consider rival causes! How else might one explain these group differences?
First, let's outline the reasoning:
CONCLUSION: (Researchers' hypothesis) Playing violent video games appears to cause an increase in violence among children.
REASON: (Researchers' evidence) Research study showed children who played violent video games were more likely to get into physical altercations with other children than were those children who did not play violent video games.Note that the words appears to cause in the conclusion tell us the researchers are making a causal claim about the evidence. But other hypotheses can explain this evidence.
The report fails to tell us how the children were selected into the three different groups. It is possible that the children were allowed to self-select what games they play, and perhaps children who are more likely to be violent tend to choose "violent" video games. If so, it is possible the researcher has the causal link reversed. Also, nothing is revealed as to how the "play" situation was set up for the children. Perhaps the room or selection of toys was set up in such a way to encourage physical altercations among those who played the
"violent" video games. We bet you can think of other reasons these groups— violent video game players and not—differ in their likelihood of resorting to physical violence.
We cannot make you aware of all possible rival causes. In the following selections, however, we provide several clues for finding common rival causes. Rival Causes for Differences Between Groups One of the most common ways for researchers to try to find a cause for some event is to compare groups. For example, you will frequently encounter the following kinds of references to group comparisons:
Researchers compared an experimental group to a control group.
One group received treatment X; the other group didn't.
A group with learning disabilities and a group without learning disabilities.
When researchers find differences between groups, they often conclude, "Those differences support our hypothesis." For example, a researcher might compare a group of people trying to lose weight treated with a new drug with a control group of people trying to lose weight that does not get the new drug, find that the groups differ in their weight loss, and then conclude that the drug caused the difference. The problem is that research groups almost always differ in more than one important way, and thus group differences often are consistent with multiple causes. Thus, when you see communicators use findings of differences between groups to support one cause, always ask, "Are there rival causes that might also explain the differences in the groups?"
Let's take a look at a study that compares groups and try to detect rival causes.
In a recent research study, students who prepare for a standardized test by taking a special course designed to teach students how to take the test have scored higher than students who prepare for the same standardized test by reviewing several books about the test.
Here we have two groups: the students who take the class and the students who read a few books. The question we need to ask is, "Did these two groups differ in important ways other than the test preparation they experienced?" Did you think of either of the following possible important differences between the two groups that might account for test score differences?
• Differences in students ' academic (and economic) background. It is possible that the course costs a substantial sum of money, and only those students who had the money could afford to take the class. Moreover, it is also possible that those students who could afford the money for the class also could afford better private school education before taking the test, and thus start off from a privileged position in comparison with the students who did not take the class.
• Differences in motivation. Perhaps the students who signed up for the class are the students who really want to excel in the test. Students who read a few books might be less interested in scoring really well on the standardized test. Alternatively, the students might have chosen study methods based on how they best learn. It is possible that those who learn best in a class setting might be predisposed to do well on standardized tests.
You probably came up with other important differences.
Many factors can cause research groups to differ!
Confusing Causation with Association
We have an inherent tendency to "see" events that are associated, or that "go together," as events that cause one another. That is, we conclude that because characteristic X (e.g., amount of energy bars consumed) is associated with characteristic Y (e.g., performance in an athletic event), that X therefore causes Y. The following are examples of such reasoning:
1. Classes with larger numbers of students enrolled tend to experience high rates of students' skipping class.
2. More red cars than any other color are pulled over for speeding; therefore, the color of the car affects how fast it goes.
When we think this way, we are, however, often very wrong! Why? Usually multiple hypotheses can explain why X and Y "go together." In fact, there are at least four different kinds of hypotheses to account for any such relationship.
Knowing what these are will help you discover rival causes. Let's illustrate each of the four with a research example.
A recent study reported that "smoking combats the flu." The researchers studied 525 smokers and found that 67 percent of the smokers did not have the flu once over the last three years. The researchers hypothesized that the nicotine in the smoke from cigarettes destroys the flu virus before it can spread and cause sickness.
Should people who are feeling under the weather run out and start smoking to prevent the onset of the flu?
Not yet. Before they do, they should contemplate each of four potential explanations for the research findings.
Explanation 1: Xis a cause ofY. (Smoking does indeed kill the flu virus.)
Explanation 2: Y is a cause ofX. (Feeling healthy, or feeling the beginning of what might be the flu, causes people to smoke.)
Explanation 3: Xand Y are associated because of some third factor, Z. (Smoking and being without the flu are both caused by related factors, such as frequent washing of the hands after smoking prevents the spread of the flu virus.)
Explanation 4: X and Y influence each other. (People who do not usually catch the flu have a tendency to smoke, and the smoke may affect some potential illnesses.)
Remember: Association or correlation does not prove causation!
Yet much evidence to prove causation is only based on association or correlation.
When an author supports a hypothesis by pointing to an association between characteristics, always ask, "Are there other causes that explain the association?"
Test yourself on the following:
A recent study reported that "ice cream causes crime." The researchers studied ice cream sales and crime rates over the last five years in the ten largest U.S. cities and found that as ice cream sales increase, so does the crime rate. The researchers hypothesized that the consumption of ice cream triggers a chemical reaction in
one's brain causing the individual to have an inclination toward crime.
We hope you can now see that people who eat ice cream need not be concerned that they are about to commit a crime. What rival causes did you think of? Couldn't the increased summer heat account for the association between ice cream sales (X) and crime (Y)?
This confusion between correlation and causation is as understandable as it is dangerous. A cause will indeed precede its effect. But many things preceded that effect. Most of them were not causal.
You should now be able to identify two common causal reasoning fallacies by attending to the above four possible explanations of why events might be associated:
F: Confusion of Cause and Effect: Confusing the cause with the effect of an event or failing to recognize that the two events may be influencing each other.
F: Neglect of a Common Cause: Failure to recognize that two events may be related because of the effects of a common third factor.
Confusing "After this" with "Because of this"
Shortly after the 2004 Summer Olympics where Michael Phelps won six gold medals and two bronze medals, the price of college tuition nationwide once again increased. Does this mean we can attribute the price of college tuition to Michael Phelps' Olympic success? No. There are many other possible causes.
If we were to infer such a conclusion, we would be illustrating a very common way that people confuse causation with association. Often, we try to explain a particular event as follows: Because event B followed event A, then event A caused event B. Such reasoning occurs because human beings have a strong tendency to believe that if two events occur close together in time, the first one must have caused the second one. To appreciate the flaw in this reasoning, pick up today's newspaper and make a list of what is going on in the world. Then pick up yesterday's newspaper and make a similar list. Could you conclude that the events of yesterday are causing the events of today? Clearly not. For example, yesterday's news contained more stories about the war in Iraq, and today's news reported that IBM is attempting to buy a start up firm to advance open-source software. It is highly unlikely that events in Iraq caused IBM's attempt to expand its business.
Many events that occur after other events in time are not caused by the preceding events. When we wrongly conclude that the first event causes the second because it preceded it, we commit the Post hoc, ergo propter hoc (meaning: "after this, therefore because of this") fallacy, or, for short, the Post hoc fallacy. Such reasoning is responsible for many superstitious beliefs.
For example, you may have written an excellent paper while wearing a particular hat, so now you always insist on wearing the same hat when you write papers.
F: Post hoc Fallacy: Assuming that a particular event, B, is caused by another event, A, simply because B follows A in time.
The following examples further illustrate the problem with this kind of reasoning.
"The quarter I found yesterday must be lucky. Since I have found it I got an A on a really hard test, my least favorite class was canceled, and my favorite movie was on TV last night." (Never mind the fact that I studied really hard for my test, my professor has a six-year old who recently had the flu, and the TV schedule is
made far in advance of my finding a quarter.)
"Ever since September 11th, 2001, large numbers of people have been afraid to fly, and airlines are suffering financially because of it." (But perhaps it is also relevant that the economy was suffering before, and continued to suffer after September 11, and the weak economy could mean that people have less disposable income with which to purchase airplane tickets.)
As you might guess, political and business leaders are fond of using the Post hoc argument, especially when it works in their favor. For example, they tend to take credit for anything good that takes place after they assumed their leadership role and to place blame elsewhere for anything bad that happens.
Remember: The finding that one event follows another in time does not by itself prove causation; it may be only a coincidence. When you see such reasoning, always ask yourself, "Are there rival causes that could account for the event?" and, "Is there any good evidence other than the fact that one event followed the other event in time?"
Explaining Individual Events or Acts
Why did Mount St. Helen's erupt again in 2004? What caused the increase in alcohol prices? Why did Martha Stuart engage in insider trading?
Like our question about Virginia Woolf s suicide, these questions seek explanations of individual historical events. Scientific research studies cannot answer the questions. Instead, we must search the past for clues. Such a search makes us highly susceptible to reasoning errors for several reasons. A few of these are especially important to remember.
First, as we saw in Virginia Woolf s case, so many different stories for the same event can "make sense." Second, the way we explain events is greatly influenced by social and political forces, as well as by individual psychological forces.
For example, men view the cause of drug abuse differently than women, and Democrats might view the causes of poverty differently from Republicans. Also, a common bias is "the fundamental attribution error," in which we typically overestimate the importance of personal tendencies relative to situational factors in interpreting the behavior of others. That is, we tend to see the cause of other's behavior as coming from within (their personal characteristics) rather than from without (situational forces.) So, for example, when someone
steals something from someone else, we are likely to view the stealing initially as a result of a tendency of the person to be immoral or to be inconsiderate.
However, we should also consider the role of outside circumstances, such as poverty or an honest mistake.
Another kind of common psychological error is to start with a limited number of possible causes and then to interpret additional information (even if it is irrelevant) as corroborating these existing hypotheses, rather than keeping the information separate or generating new, perhaps more complex, hypotheses.
Our tendency is to simplify the world; yet often explanations require much complexity. Explaining events is not as simple as frequently portrayed by guest experts on the popular talk shows. How can we know whether we have a "good" explanation of a particular event or set of events? We can never know for sure. But we can make some progress by asking critical questions.
Be wary of accepting the first interpretation of an event you encounter. Search for rival causes and try to compare their credibility. We must accept the fact that many events do not have a simple explanation.
Evaluating Rival Causes
The more plausible the rival causes that you come up with, the less faith you can have in the initial explanation offered, at least until further evidence has been considered. As a critical thinker, you will want to assess as best you can how each of the alternative explanations fits the available evidence, trying to be sensitive to your personal biases.
In comparing causes, we suggest that you apply the following criteria:
1. their logical soundness;
2. their consistency with other knowledge that you have; and
3. their previous success in explaining or predicting events.
USING THIS CRITICAL QUESTION
Every assertion about causation should trigger immediate curiosity in the mind of a critical thinker. But are there alternative causes for the phenomenon?
Asking someone to consider rival causes is constructive criticism at its finest. Your objective in doing so is to find a better causal explanation.
Evidence and Your Own Writing and Speaking
The last three chapters have indirectly provided you with a clue for effective communication. Your audience will be justifiably impressed when you provide strong and sufficient evidence for your claims. But implicit in this clue is a warning: Your audience expects and should demand that your claims are supported
by thorough evidence. Satisfying this demand is one of your greatest challenges as a writer and speaker.
Factual claims about the causes of events are weakened when other claims about the causes can be offered. Such claims are rival causes. A common logical error in explaining observations is to confuse causation
with association. Thus, always ask what other causes might explain observed associations. Be especially alert to the Post hoc fallacy.
Critical Question: Are there rival causes?
Each of the following examples provides an argument to support a causal claim. Try to generate rival causes for such claims. Then try to determine how much you have weakened the author's claim by knowledge of rival
Oranges to combat the blues. Researchers have recently revealed that eating two oranges a day can help alleviate depression. Researchers studied 13 patients who had feelings of depression. After three weeks of eating two oranges a day, 9 of the 13 people reported improvement in their condition. The researchers hypothesize that the citric acid and vitamin C in oranges helps to stimulate serotonin production, helping to combat depression.
Why did the corporate executive steal funds from his business? A close look at his life can provide a clear and convincing answer. The executive comes from a very successful family where his parents are doctors and his siblings are lawyers.
As a corporate executive, he was not making as much money as his family members. Also, the executive believes heavily in the American dream and the idea that if one works hard enough that person will succeed. However, despite his hard work the executive has had a number of recent business failures, including losing a substantial sum of money in the stock market. To make matters worse, his children need braces. To live up to expectations, become a success, and provide for his family, the executive had to steal the money from his business.
According to a recent study, one of the major causes of violence in schools is listening to aggressive heavy metal music. Researches studied more than 100 cases of "serious" violence within schools, and have found that 68% of the children involved in the violence listened to heavy metal music. These children would frequently come to school with headphones, listening to this music, as well as wearing clothing from heavy metal bands. Frequently these heavy metal songs discuss violence, and therefore are a direct cause of school violence.
CONCLUSION: Eating oranges helps alleviate depression.
REASON: 9 of 1 3 patients who ate oranges experienced improvement with their depression.
Can anything else account for the change besides eating oranges? Yes; the researcher fails to rule out many obvious alternative explanations. For example, the patients might have expected to get better, and these expectancies might have led to feeling better. Also, they knew the purpose of eating oranges, and a rival cause is that they tried to please the researchers by reporting that they felt better. We can also hypothesize that external events during the three-week treatment period caused the change. Perhaps during the three weeks of treatment, for example, the weather was especially good, and these people spent much more time exercising outside than usual, which could also help alleviate depression. Another possibility is that these people were suffering from a form of depression from which they could naturally expect to recover in a short
period of time. Can you locate other rival causes?
CONCLUSION: The executive stole money from his company to compete with his family members, to show that he is not a failure, and to provide for his family.
REASON: The executive was probably concerned with all of the above elements.
It is possible that all of the above factors were important in causing the corporate executive to steal from his company. But many other people in society have the same pressures put upon them and they do not resort to illegal means to obtain money. Are there other possible causes for such behavior? As in the case of Virginia Woolf, we suspect there may be many other plausible explanations. Before we could conclude that these stresses in the executive's life are the causal factors, we would want to know more about his childhood and more about recent events in his life. For example, has the corporate executive had any recent disagreements with his boss? Had he been using drugs? Had he had any recent highly stressful experiences? Did he have a history of stealing? After the fact, we can always find childhood experiences that make sense as causes of
adult behavior. Before we draw causal conclusions, however, we must seek more evidence to prove that the one set of events caused the other than the mere fact that one set of events preceded the other set. We must also be wary not to fall victims to the fundamental attribution error and be certain to consider external
causal factors, as well as internal ones.
Are There Rival Causes?
While an author might offer an explanation for why certain events occurred, other explanations might be plausible. When you try to identify rival causes, you are finding alternative explanations for an event. If you can identify alternative explanations, you must decide whether you should believe the author's explanation
or one of the other explanations. If the author does not provide reasons for why you should accept her explanation over other explanations, you should not be willing to accept her explanation and, ultimately, her conclusion. Thus, looking for rival causes is another step in deciding whether to accept or reject an argument.