I’ve had a lifelong love affair with Scientific American. I began reading it as a freshman in high school, particularly Martin Gardner’s monthly column on “Mathematical Games,” and I have subscribed to the magazine off and on throughout my adult life. Even though I have felt that over the years the articles have gotten less accessible to the lay reader, perhaps that’s more a function of my waning patience than it is the actual difficulty level of the articles.
In the most recent issue there is an article by two psychologists, entitled “The World Without Free Will.” I thought, aha, here finally will be scientific evidence to support my position that free will does not exist. I saved the article for last, thinking how much I would savor what the authors had to say. I was in for a major disappointment, but not necessarily for the reasons you might think.
The authors did not attempt to address the question of whether free will exists. They did say that “recent neuroscience studies have added fuel to . . . the notion that human behavior results from a complex sequence of cause and effect that is completely out of our control. The universe simply does not allow for free will.” But they also stated that “not everyone agrees, of course, and the debates over the existence of free will continue to rage.” So much for that.
What the authors were interested in was whether a person’s level of belief in free will would have an impact on his or her other behaviors or beliefs. It’s an interesting question. Specifically, they wondered if those who believe in free will (call them free will believers or FWBs) would have a different attitude toward personal responsibility than those who are skeptical of free will (call them free will skeptics or FWSs). They also wondered if FWSs would be more willing to flout rules of conduct, such as cheating or hurting others. Finally, they wondered if FWSs would have a different approach to punishment for wrongdoing than would FWBs. And in this article the authors reported on experimental studies that addressed these questions.
In brief, the authors reported that FWSs were much more likely to cheat on a test than were FWBs. Moreover, FWSs were more likely to administer gratuitous pain to others than were FWBs. The implication is that belief in free will may be important to preserving the notion of personal responsibility and that without such notion an individual is more apt to break the rules. Taken to the extreme, this would seem to support the idea that loss of belief in free will would jeopardize societal and cultural standards. As a confirmed free will skeptic, I was disappointed in these conclusions, because I have always felt that, despite the questions of personal responsibility that are raised in the absence of free will, I can and should (perhaps “must” is a better word?) live a moral life.
Having taught experimental social psychology at the college level in the far distant past, I still have an interest in the methodology that researchers in that field employ to obtain their results. In this article the authors stated that they had reduced the beliefs of the experimental subjects in free will by having them read a passage. Really? What passage, I wondered, could the subjects have read that would have any sort of immediate impact on such a fundamental belief as whether free will exists. In my personal experience, I don’t seem to have been able to change anyone’s belief regarding free will despite my best efforts at persuasive discourse.
In attempting to track down the experimental design details, I found that in the Scientific American article the authors did not specifically cite any of the studies that their conclusions were based on, though they did reference a couple of books they had written on the subject. Even so, I was able to track down one of the primary studies that they relied on by rooting around on Google. That study can be found here. The article, which was actually conducted in 2003 and published in 2008, reported that the test subjects were asked to read a passage from The Astonishing Hypothesis, by Francis Crick, a book that argues, inter alia, that free will is an illusion. Control subjects were asked to read a different passage from the Crick book that dealt with consciousness but not with the subject of free will. All subjects were then asked to complete a questionnaire designed to identify the extent to which they supported the idea of free will. Following that, they were asked to complete a test of mental math skills and were, in effect, given the opportunity to cheat or not cheat on their answers but were told not to cheat. Hmm . . . . According to the research article, those who read the anti-free-will passage had significantly lower scores on their belief in free will and cheated 50% more than those who read a neutral passage. Wow.
Importantly, when searching for these articles, I came upon a blog article by another researcher, Rolf Zwaan. Zwaan attempted to replicate the cheating study but was unable to do so. In Zwaan’s study reading the anti-free-will passage did not result in lower free will belief scores nor did it result in more cheating. There were differences in experimental design between the two studies that are worth mentioning. First, Zwaan used 150 subjects versus only 30 used in the Vohs study. Second, the subjects in the original experiment were all students at the University of Utah and half were Mormons who admitted that they had a strong connection to their faith. Zwaan’s subjects were much more diverse, both in terms of age and education as well as in religious affiliation. Finally, Zwaan questioned the extent to which the subjects bought into the premise that the experimenters would not be able to detect any cheating. And, frankly, the entire set-up of the original study seemed contrived to me. The students knew they were in a psychology experiment and must have wondered what could be the point of an experiment that involved reading a passage about consciousness and/or free will, answering questions about free will, and solving mental math problems where cheating was possible. Wouldn’t you be suspicious and wonder what the experimenters were expecting of you?
I feel there are two main points to be learned from this experience. First, the fact that an article appears in a “prestigious” journal doesn’t mean one should accept the results at face value. I had always felt that I should have more confidence in an article appearing in Scientific American than in one appearing in, say, Psychology Today. Now I’m not so sure. I know a little about social psychology research but next to nothing about biological or physics research. What am I to think? Second, If results seem too unusual (the “woo-woo” factor), one should ask questions. Extraordinary results require extraordinary—and well documented—evidence. In short, be skeptical.
© 2014 John M. Phillips