{33} V
Arguments about Causes
Did you know that students who sit at the front of the classroom tend to get better grades? And that people who are married are, on average, happier than people who aren’t? Wealth, by contrast, doesn’t seem to correlate with happiness at all—so maybe it is true after all that “the best things in life are free.” If you’d rather have the money anyway, you might be interested to know that people with “can-do” attitudes tend to be wealthier. So you’d better work on your attitude, eh?
Here we come to arguments about causes and their effects—about what causes what. Such arguments are often vital. Good effects we want to increase, bad effects we want to prevent, and we often want to give appropriate credit or blame for both. It won’t surprise you, though, that reasoning about causes also takes care and critical thinking.
The evidence for a claim about causes is usually a correlation—a regular association—between two events or kinds of events: between your grades in a class and where you sit in the classroom; between being married and being happy; between the unemployment rate and the crime rate, etc. The general form of the argument therefore is:
Event or condition E1 is regularly associated with event or condition E2.
Therefore, event or condition E1 causes event or condition E2.
{34} That is, because E1 is regularly associated with E2 in this way, we conclude that E1 causes E2. For example:
People who meditate tend to be calmer.
Therefore, meditation calms you down.
Trends may also be correlated, as when we note that increasing violence on television correlates with increasing violence in the real world.
Shows on television portray more and more violence, callousness, and depravity—and society is becoming more and more violent, callous, and depraved.
Therefore, television is ruining our morals.
Inverse correlations (that is, where an increase in one factor correlates to a decrease in another) may suggest causality too. For example, some studies correlate increased vitamin use with decreased health, suggesting that vitamins may (sometimes) be harmful. In the same way, noncorrelation may imply lack of cause, as when we discover that happiness and wealth are not correlated and therefore conclude that money does not bring happiness.
Exploring correlations is also a scientific research strategy. What causes lightning? Why do some people become insomniacs, or geniuses, or Republicans? And isn’t there some way (please?) to prevent colds? Researchers look for correlates to these conditions of interest: that is, for other conditions or events that are regularly associated with lightning or genius or colds, for example, but without which lightning or genius or colds don’t tend to happen. These correlates may be subtle and complex, but finding them is often possible nonetheless—and then (hopefully) we have a handle on causes.
Arguments from correlation to cause are often compelling. However, there is also a systematic difficulty with any such claim. The problem is simply that any correlation may be {35} explained in multiple ways. It’s often not clear from the correlation itself how best to interpret the underlying causes.
First, some correlations may simply be coincidental. For example, though the Seattle Seahawks and the Denver Broncos both went to the Super Bowl in the same year that their home states legalized marijuana—2012—it’s not likely that these events were actually connected.
Second, even when there really is a connection, correlation by itself does not establish the direction of the connection. If E1 is correlated with E2, E1 may cause E2—but E2 may instead cause E1. For example, while it is true (on average) that people with “can-do” attitudes tend to be wealthier, it’s not at all clear that the attitude leads to the wealth. It may be more plausible the other way around: that the wealth causes the attitude. You’re more apt to believe in the possibility of success when you’ve already been successful. Wealth and attitude may correlate, then, but if you want to get wealthier, just working on your attitude may not get you very far.
Likewise, it’s entirely possible that calmer people tend to be drawn to meditation, rather than becoming calmer because they meditate. And the very same correlation that suggests that television is “ruining our morals” could instead suggest that our morals are ruining television (that is, that rising real-world violence is leading to an increase in the portrayal of violence on television).
Third, some other cause may underlie and explain both of the correlates. Again E1 may be correlated with E2, but rather than E1 causing E2 or E2 causing E1, something else—some E3—may cause both E1 and E2. For example, the fact that students who sit in the front of the classroom tend to get better grades may not imply either that sitting in the front leads to better grades or that getting better grades leads to sitting in the front of the class. More likely, some students’ special commitment to making the most of their schooling leads both to sitting in the front of the classroom and to better grades.
Finally, multiple or complex causes may be at work, and they may move in many directions at the same time. Violence on television, for example, surely reflects a more violent state of society, but also, to some degree, it surely helps to worsen that violence. Quite likely there are other underlying causes as well, such as the breakup of traditional value systems and the absence of constructive pastimes.
{36}
Since a variety of explanations for a correlation are usually possible, the challenge for a good correlation-based argument is to find the most likely explanation.
First, fill in the connections. That is, spell out how each possible explanation could make sense.
NO:
Independent filmmakers generally make more creative films than the big studios. Thus, their independence leads to their creativity.
There’s a correlation, yes, but the causal conclusion is a little abrupt. What’s the connection?
YES:
Independent filmmakers generally make more creative films than the big studios. It makes sense that with less studio control, independent filmmakers are freer to try new things for more varied audiences. Independents also usually have much less money at stake, and therefore can afford for a creative experiment to fall flat. Thus, their independence leads to their creativity.
Next, try to fill in the connections in this way not just for the explanation you favor, but also for alternative explanations. For example, consider studies that correlate increased vitamin use with decreased health. One possible explanation is that vitamins actually worsen health, or anyway that some vitamins (or taking a lot of them) are bad for some people. It is also possible, though, that people who already are in bad or worsening health may be using more vitamins to try to get better. In fact, this alternative explanation seems, at least at first glance, equally or even more plausible.
{37} Finally, try to decide which is the most likely explanation for the correlation. You may need more information. In particular, is there other evidence that (some?) vitamins can sometimes be harmful? If so, how widespread might these harms be? If there is little direct and specific evidence of harm to be found, especially when vitamins are taken in appropriate dosages, then it’s more likely that poorer health leads to more vitamin use than that more vitamin use leads to poorer health.
Or again: Marriage and happiness correlate (again, on average), but is it because marriage makes you happier or because happier people tend to be more successful at getting and staying married? Fill in the connections for both explanations and then step back to think.
Marriage clearly offers companionship and support, which could explain how marriage might make you happier. Conversely, it may be that happy people are better at getting and staying married. To me, though, this second explanation seems less likely. Happiness may make you a more appealing partner, but then again it may not—it could instead make you more self-absorbed—and it is not clear that happiness by itself makes you any more committed or responsive a partner. I’d prefer the first explanation.
Note that the most likely explanation is very seldom some sort of conspiracy or supernatural intervention. It is possible, of course, that the Bermuda Triangle really is spooked and that is why ships and planes disappear there. But that explanation is far less likely than another simple and natural explanation: that the Bermuda Triangle is one of the world’s heaviest-traveled shipping and sailing areas, with tropical weather that is unpredictable and sometimes severe. Besides, people do tend to embellish spooky stories, so some of the more lurid accounts, having passed through countless retellings, aren’t (let’s just say) the most reliable.
Likewise, although people fasten onto inconsistencies and oddities in dramatic events (the JFK assassination, 9/11, etc.) to justify conspiracy theories, such explanations usually leave a great deal more unexplained than the usual explanations, however incomplete. (For instance, why would any plausible conspiracy take this particular form?) Don’t assume that every little oddity must have some nefarious explanation. It’s hard enough to get the basics right. Neither you nor anyone else needs to have an answer for everything.
{38}
Plenty of happy people are not married, of course, and plenty of married people are unhappy. Still, it does not follow that marriage has no effect on happiness on average. It’s just that happiness and unhappiness (and, for that matter, being married or unmarried) have many other causes too. One correlation is not the whole story. The question in such cases is about the relative weight of different causes.
If you or someone else has argued that some E1 causes some E2, it is not necessarily a counterexample if occasionally E1 doesn’t produce E2, or if another cause entirely may also sometimes produce E2. The claim is just that E1 often or usually produces E2, and that other causes less commonly do, or that E1 is among the major contributors to E2, though the full story may involve multiple causes and there may be other major contributors too. There are people who never smoke cigarettes at all and still get lung cancer, and also people who smoke three packs of cigarettes a day and never get it. Both effects are medically intriguing and important, but the fact remains that smoking is the prime cause of lung cancer.
Many different causes may contribute to an overall effect. Though the causes of global climate change are many and varied, for instance, the fact that some of them are natural, such as changes in the sun’s brightness, does not show that human contributions therefore have no effect. Once again, the causal story is complex. Many factors are at work. (Indeed, if the sun is also contributing to global warming, there’s even more reason to try to decrease our contribution.)
Causes and effects may “loop,” too. Filmmakers’ independence may lead to their creativity, but, then again, creative filmmakers may seek independence from the start, leading to more creativity, and so on. Others may seek both creativity and independence because they prefer a less pressured life, or maybe they just have some great idea that they can’t sell to a big studio. It’s complicated….