A Psychologist's Unconvincing Explanation for Awe

One of the most sublime human emotions is the one we call awe. It's not very common for a modern person to feel it. But imagine you are a New York City dweller used to seeing maybe two or three stars in the sky. Imagine you take a vacation in Colorado. You book a campsite in Rocky Mountain National Park. After nightfall you lie near your tent and look up at the sky. You are astonished. You can now see not just two or three stars, but thousands of stars. Plus you can see some strange faint band stretching across the sky. It looks like some ghostly river. You realize you are looking at the plane of the Milky Way galaxy. You suddenly feel a strange emotion you have rarely felt before. It is as if you have got in touch with some magnificent reality vastly greater than your little self. You experience an awe you will long remember. 

But why do people even feel such a rare emotion? In last week's edition of the Huffington Post, psychology professor Dacher Keltner attempts an explanation, in a rather long article entitled Why Do We Feel Awe? But his explanation doesn't hold water. He starts off with this suggestion:

Why did awe became part of our species’ emotional repertoire during seven million years of hominid evolution? A preliminary answer is that awe binds us to social collectives and enables us to act in more collaborative ways that enable strong groups, thus improving our odds for survival.

This hypothesis is unbelievable. Awe has nothing us to do with binding to social collectives, nothing to do with enabling social groups, and nothing to do with collaboration. Awe does nothing to improve any organism's odds for survival.

We might be able to explain fear using a Darwinian explanation, on the grounds that an organism that is afraid of scary sights is more likely to flee predators. But awe is something different from fear. When you look up at a sky filled with stars, you feel awe, but you feel no fear at all. Nothing could be less scary than a distant star.

To try to justify his explanation for why humans feel awe, Keltner cites an experiment:

My colleague Michelle Shiota had participants fill in the blank of the following phrase: “ I AM ____.” They did so 20 times, either while standing before an awe-inspiring replica of a T. rex skeleton in UC Berkeley’s Museum of Paleontology or in the exact same place but oriented to look down a hallway, away from the T. rex. Those looking at the dinosaur were more likely to define their individual selves in collectivist terms—as a member of a culture, a species, a university, a moral cause. Awe embeds the individual self in a social identity.

This is rather hilarious. The skeleton was not even a real T. rex skeleton, but only a replica (probably something made out of plaster or plastic). Why would someone feel awe looking at some fake dinosaur skeleton? 

Equally questionable is the experiment which Keltner describes below, involving some trees near a campus science building:

Participants first either looked up into the tall trees for one minute—long enough for them to report being filled with awe—or oriented 90 degrees away to look up at the facade of a large science building. They then encountered a person who stumbled, dropping a handful of pens into the dirt. Sure enough, the participants who had been gazing up at the awe-inspiring trees picked up more pens.

This is as dubious as the experiment with the fake dinosaur skeleton. Looking up at tall trees doesn't produce awe. Trees are too common to produce a feeling of awe. The fact that some participants may have reported that they felt awe (when presented with a questionnaire asking whether they did) is probably just a case of suggestibility or people reporting that they had a feeling that they thought they were supposed to have. Similarly, if you show a picture of a beggar to a man and ask if it makes him sad, someone who does not feel sad will often say he does feel sad, as a kind of act of social obligation or acting in the suggested or expected way. Since Keltner merely says that these tree gazers “picked up more pens” rather than saying they picked up “many more pens” (and since the paper linked to does not mention any specific numbers regarding this experiment or any level of statistical significance), we can assume the effect he is reporting is minimal or perhaps not even statistically significant. Such a result tells us nothing.

At this link the experiment is described in detail, and the paper claims that “Participants who gazed up at the trees offered more help to an experimenter than did participants who gazed up at a building,” but offers no specific numbers backing up such a claim. So we must conclude the effect was minimal or marginal – for all we know, it could have been merely a “1% greater” type of effect.

What we have here is rather silly psychology experimentation on a shoestring budget. To do a decent experiment on awe, you should do something like take people to the Grand Canyon or to a mountain place with crystal clear air where you can see 6000 stars at night. Then when you suddenly showed them the awesome scenery, they might experience awe. The experiment Keltner describes are “science on a shoestring” type of experiments that probably tell us nothing about awe, because they don't involve things that produce awe to a significant degree. It's as if the experimenters were too lazy to leave their local campus, and find something really awe-inspiring.

I may note that inexpensive short-duration experiments like this are generally of little worth whenever they report modest effects or borderline effects (as in this case). If some college does a 2-year long study costing 5 million dollars, that has some weight, because presumably there would not been have time and money to try such a study multiple times and then report only one version. But it's a totally different situation for inexpensive short-duration studies. Let's say I'm a professor trying to show that wearing some color of shirt affects your test score performance. I could do 20 one-day studies (asking my students to wear a particular color on each of 20 test days), and then cherry pick a particular day, which ever day seemed to best support a “shirt color influences test scores” hypothesis. I could then author a scientific paper reporting only on that particular day's test. Of course, that really wouldn't give any evidence for such a hypothesis. I would just be making an inappropriate use of random fluctuations in test data.

Also of little evidence value is a TV watching study described in this paper (Study 3), which reports only marginal results. On page 8 of the pdf, the authors report that they tried to experimentally induce awe by showing a “5-minute clip inducing awe, consisting of nature clips from the BBC’s Planet Earth series composed of grand, sweeping shots of scenic vistas, mountains, plains, forests, and canyons.” Such “eye candy” clips don't actually produce awe. If you see a real canyon, it may produce awe, but seeing one on television will not (unless you've never seen a canyon before on television, or unless you're watching a good science fiction showing some type of stunning scenery you've never seen before). Moreover, the procedure described in Study 3 is so convoluted that it is lacking in any evidence value. A supposed slight increase in generosity was measured by a willingness to donate points in some computer game, but the supposed difference in generosity involved giving away a few more points which each had a cash value of only pennies.

Similarly weak from an evidence standpoint is Study 1 in the paper. Based on a very dubious analysis of a person's tendency to feel awe, the study reports a weak .123 correlation between awe and a tendency to give away imaginary money in a game. I need not say much about the weakness of that, other than to point out that a compelling correlation is one that is, say, something like .700. Even when correlations are much greater than .700 they are often coincidental. This web site lists correlations of greater than .900 between totally unrelated things, such as a .9925 correlation between the divorce rate in Maine and the per capita consumption of margarine.

I find the experiments of Keltner and his colleagues on this matter to be quite unconvincing. It seems these are the type of results that anyone could get to support any random psychology hypothesis he wanted to support, just by doing 20 shoestring-budget short-duration experiments and then reporting the few in which random data variations best supported the hypothesis.

Keltner tries to suggest that awe is something very social, but it isn't. Quite to the contrary, awe is the least social of all emotions. When you look at a sky filled with stars, you are absorbed in that external glory, and are least likely to be thinking about another human.

Keltner's suggestion that awe is something that improved an early human's odds of survival is without any merit. Quite to the contrary, we should assume that awe is something that decreased an early human's odds of survival. Show me a caveman who spent quite a bit of time staring up at the stars with a feeling of awe, and I will show you a caveman more likely to have been killed by a predator while he is distracted by this activity that did nothing to help his chances of survival. Show me a caveman who spent quite a bit of time watching the sun set while he felt awe, and I will show you a caveman who would have been more likely to be killed by a predator while he is distracted by this activity that did nothing to help his chances of survival. Show me a caveman who tended to go out of his cave and watch a lightning storm with awe, and I'll show you a caveman more likely to have been struck dead by a lightning bolt. Show me a caveman who stood staring in awe at the big tusks of a mastadon, and I'll show you a caveman more likely to have been gored to death by those tusks. Show me a caveman fond of climbing mountains to experience awe-inspiring vistas, and I'll show you a caveman more likely to die in an accident while scaling such heights.

There is no plausible Darwinian explanation for the emotion of awe, just as there is no plausible Darwinian explanation for numerous other aspects of the human mind – things such as musical ability, grammar ability, philosophical reasoning, spirituality, insight, altruism, and mathematical ability. As I argue here, these are things that do not increase an organism's survival value in a natural setting, and which therefore cannot be explained through a simplistic explanation of natural selection. We must postulate that something much more was involved in the origin of humanity than just random mutations and natural selection.

If you have a spark of the divine in you, or a soul, you should probably expect that encountering some reality much grander than yourself might trigger some sublime emotion, something like awe. But if you were merely the soulless product of blind chance, you should expect no such thing. If blind Darwinian processes were all that were involved in our origins, you should expect that when you look up at a mountain sky of 6000 dim, distant stars, you should feel absolutely nothing.