Linda: A prelude
Meet Linda ->
Linda is 31 years old, single, outspoken, and very bright. She majored in philosophy. As a student, she was deeply concerned with issues of discrimination and social justice, and also participated in anti-nuclear demonstrations.
Now that you know a bit about Linda, here's a question for you...
Which is more probable?
- Linda is a bank teller
- Linda is a bank teller and is active in the feminist movement.
If you're like most people, you picked Option 2 (or were at least sorely tempted to). But that is incorrect. Why, you ask? Because feminist bank tellers are just a subset of bank tellers in general. Option 1 is always more probable.
Tversky and Kahneman famously used this problem to demonstrate the existence of the representativeness heuristic. People intuitively (and incorrectly) guess Option 2 because the contents of the description (liberal, etc) just sounds like the group (feminists) implied by Option 2. This leads people to jump to the conclusion that Option 2 is more probable. Go with your gut, and you get this one wrong.
But you can break this effect. Consider if I gave you the same description of Linda, but asked instead...
Which is more probable?
- Linda is a bank teller
- Linda is a bank teller and is an avid big game hunter
Now Option 2 just doesn't feel right. Basically, in this example, there's no intuitive linkage between the description and the implied group membership. In a pilot study, upwards of 80% of people picked Option 2 if it implies that Linda might be a feminist, but 0% of people picked Option 2 if it implies Linda shoots big critters for fun.
I know what you're thinking now: "That's kind of neat, but what does that have to do with having sex with dead chickens?"
To which I would respond: "Settle down, pervert, we'll get to the necrobestiality soon enough." The key point of that opening was that people only tend to pick Option 2 in a representativeness heuristic task when the contents of the description are intuitively seen as congruent with the group membership implied by Option 2.
In a recently published paper, I decided to play around a bit with the representativeness heuristic in order to gauge people's intuitions about a link between religion and morality.
And what, you ask, is the link between religion and morality?
First off, let's ask a scientist. And the scientist will likely tell you that question actually has a complicated answer. On the one hand, certain types of religious beliefs seem to promote within group cooperation, aid in the development of stable communities, and do other nice things like help people resist temptation. That sounds good! On the other hand, we also know that core moral intuitions are present in preverbal (hence, presumably, pre-religious) infants, as well as in nonhuman primates. At least some of our moral intuitions seem to be robust across differences in religion. So maybe religion does something with morality...but it certainly isn't doing all the heavy lifting.
There's the scientist's answer. What if we ask a nonscientist the very same question about a link between religion and morality? Well, the answer might not be nearly as nuanced. Around the globe, many if not most people report that belief in God is a necessary prerequisite for morality. No God? No good! To the extent that there are so many people and cultures that view religion as a (if not the) source of morality, this means that people might use the moral conduct of others as a cue to their religious beliefs. If you think that belief in God is the basic building block of morality and you see someone doing something good, you might intuitively assume that she is religious. But if you see someone doing something bad, you might intuitively jump to the conclusion that she is not religious.
So, here was the core idea of a bunch of studies:
Do people intuitively assume that the perpetrators of immoral acts are atheists?
And the method employed across these studies was a twist on the classic Linda problem that opened this post. Instead of giving people a description of a smart, liberal woman and seeing if people jumped to the conclusion that she was a feminist (rather than a big game hunter), I gave people descriptions of immoral acts and then systematically varied the group of Option 2 in the representativeness heuristic task to see which groups people intuitively link up to descriptions of moral transgressors.
Now, morality is a complicated beast, and people tend to have different categories of moral judgments. Excellent work by Jonathan Haidt and his colleagues suggests the operation of at least 5 core domains of morality: harm, fairness, loyalty to the ingroup, obedience to authority, and purity (respect for a predetermined "natural order"). Not everybody considers all of these categories as morally relevant (for instance, political conservatives place much more weight on authority, loyalty, and purity than do liberals), but these five make a good starting point if you want to capture the diversity of moral thought.
In one study, I grabbed a bunch of participants. Each participant would read a short vignette in which someone is described as violating one of these moral foundations (drawing heavily from previous work from the Haidt & Co. camp's work). So, for instance, a participant might read about Russell making fun of an obese woman and kicking a dog (harm), or Brad cheating at cards (fairness), or Lesley pretending to be Canadian (loyalty), or Drew telling off a bunch of cops (authority), or Catherine the medical lab tech deciding to sample a bit of meat from a human cadaver (eewwwwww, and a purity violation of gross magnitude). Then, each participant would be asked whether it is more probable that the perp was:
1) A teacher, or
2) A teacher who is XXXXXX
Participants readily picked Option 2 in all of these cases when XXXXXX was "an atheist (someone who does not believe in God)." Just as the description of Linda makes people--at a gut reaction level--think "hmmmm...sounds a bit feminist-ey" descriptions of all sorts of moral violations made people think "hmmmm....sounds a bit atheist-ey." But, participants would not pick Option 2 frequently when XXXXXX was another disliked group (gay). Here's a graph of that -->
So, if someone does something immoral--across a wide range of moral violations--people readily and intuitively assume that the person does not believe in God, but they don't assume the person is gay (gays being included as a control group...if you want the rationale for why that's a good control group read the paper).
What about other immoral acts? What about other control groups? What about the chicken already?
In a couple of other studies, I used the exact same methods to answer these questions. And I picked some classic examples of moral violations that don't necessarily cause harm to anyone, but somehow, they just seem wrong to most folks. We can't really come up with a compelling, rational reason for why we we think they're wrong, but nonetheless people tend to find these acts immoral. For example, in one study, people read this description (again modified from Haidt's stuff...blame him, not me):
Graeme and his sister were traveling together in France. One night they were staying alone in a cabin near the beach. They decided that it would be interesting and fun if they tried making love. At very least it would be a new experience for each of them. Graeme's sister was already taking birth control pills, but Graeme used a condom too, just to be safe. They both enjoyed it, but they decided not to do it again. They keep that night as a special secret between them, which makes them feel even closer to each other.
Or try this description on for size:
On the way home from work, Jack decided to stop at the butcher shop to pick up something for dinner. He decided to roast a whole chicken. He got home, unwrapped the chicken carcass, and decided to make love to it. He used a condom, and fully sterilized the carcass when he was finished. He then roasted the chicken and ate it for dinner alongside a nice glass of Chardonnay.
Now the million dollar representativeness heuristic question:
Which is more probable?
1) Jack/Graeme is an X
2) Jack/Graeme is an X who Y
And if Y is "does not believe in God," most participants pick Option 2. Evidently those descriptions just "fit" people's intuitive conception of an atheist. But people didn't pick Option 2 if Y was a whole host of religious and ethnic groups.
Finally (actually, this one is Experiment 1 in the paper...I decided to present the studies chronologically in terms of when I ran them for this post), I decided to give people a description of someone doing something that is clearly, unambiguously, incontestably immoral, just to see if I could find something soooo bad that my participants wouldn't think even an atheist would want to do it. Enter...Dax.
When Dax was young, he began inflicting harm on animals. It started with just pulling the wings off flies, but eventually progressed to torturing squirrels and stray cats in his neighborhood.
As an adult, Dax found that he did not get much thrill from harming animals, so he began hurting people instead. He has killed 5 homeless people that he abducted from poor neighborhoods in his home city. Their dismembered bodies are currently buried in his basement.
Yup, Dax is your classic psychopath serial murderer.
So, which is more probable?
1) Dax is a teacher
2) Dax is a teacher who does not believe in God
According to a bunch of my participants, Option 2 is more probable. Evidently because that description is intuitively viewed as representative of atheists. However, if Option 2 was any of five other religious groups (Christian, Jewish, Muslim, Hindu, Buddhist), people seldom picked it. Serial killer equals atheist, but serial killer does not equal any other category of religion.
Here's a graph summarizing three experiments, all showing essentially the same results, over and over again.
So, to recap, if people read a description of someone engaging in immoral behavior, they intuitively assume that the perpetrator does not believe in God. The effects were fairly specific to atheists, as they were not apparent for any of 11 other control groups. And the effects showed up across 8 different moral violations that varied greatly in scope, severity, and use of chicken. Just as many people assume that religion is a key component of morality, they also assume that without religion, people are capable of some pretty nasty business. Or, to put it differently, a lot of people seem to assume that religion is what's preventing people from incest, serial murder, disturbing culinary preferences, and pretending to be Canadian. Without religion? Sure, why not do those things?
Sounds like Dostoevsky was on to something with this line:
Without God and the future life? It means everything is permitted now, one can do anything?
-Dostoevsky, The Brothers Karamazov
So once again, I got scooped. Dostoevsky nailed the perception, and beat me to it by 134 years. But my graphs are prettier.
Right about now, you might be asking yourself some questions about who my participants were. Something like this:
It turns out that the participants were all recruited from Amazon's Mechanical Turk service. They aren't a particularly religious bunch either. I asked everyone to list their religious affiliations and to rate their strength of belief in God from 0 (God definitely doesn't exist) to 100 (God definitely exists). And here is my sample makeup across studies:
From "Dawkins" to "Devout" my participants trend more towards Dawkins. So, even though lots of my participants weren't religious, they overall still thought that immoral acts were likely performed by atheists. As a very strong and conservative test, I next grabbed only my atheist participants (defined as people who checked the "atheist" box for religious affiliation and rated their belief in God at 0). Pooling across all of the studies, even my atheist participants were still more likely to pick Option 2 when it implied an atheist identity than when it implied any of the other 11 groups.
Put all of this together, and a consistent story emerges: given a description of someone doing something immoral, people intuitively assume that the perpetrator does not believe in God. And even atheist participants shared this intuition, albeit to a smaller extent.
So, to answer the question in the title of this post, what kind of monster makes sweet love to dead chickens?
According to my participants, only an atheist monster.
Hopefully obvious and unnecessary caveat, unfortunately brought on by misinterpretations of previous research: I'm testing whether people have the perception that atheists are capable of nasty business. I'm not saying atheists actually do nasty business any more than anyone else.