Field of Science

Religious services, not belief, make you want to kill strangers

Suppose you're lost in a strange town and it's late at night and you see a group of men coming towards you. Do you feel more safe or less safe on knowing they've just come from a prayer meeting?
Christopher Hitchens was famously challenged with this question - and swiftly replied that it would be no reassurance to him at all (citing "Mr Paisley's Martyrs Memorial church in Belfast or a Party of God soiree in Beirut or to the Greater Serbia Church in downtown Belgrade" as examples). There's a gut feeling here that getting groups of people together to celebrate their religion might not have the great consequences that religious leaders like to believe it has!

But is that so? Religion has been linked with both pacifism and violence in a number of contexts. In the modern world, there's an urgent need to unravel the apparent links between suicide attacks and Islam. There's a substantial minority of opinion that says that this link is due to something special about Islam. It's supposed to be a particularly violent religion.

New research from Jeremy Ginges at The New School in New York pretty convincingly shows that religious beliefs probably have little or nothing to do with it. In fact, it's all about religious attendance.

In collaboration with Ara Norenzayan at the University of British Columbia, Ginges conducted four studies trying to figure out whether it was beliefs or participation that was most closely linked to support for suicide bombings. In the first two studies, they simply analysed data from existing surveys conducted among Palestinian Muslims. In both studies, frequency of praying wasn't linked to attitudes to the bombers. But those who attended a mosque every day were roughly twice as likely to support suicide attacks.

This is interesting, but it might just be circumstantial. So next they tried something rather more clever. They switched their focus to Jewish settlers, rather than Arabs. They phoned up a random selection of people, and asked them about their support for a specific event - the attack by Baruch Goldstein in 1994 in Hebron, which killed 29 muslims. Was this act 'extremely heroic' they asked.

Now for the clever bit. For one-third of the sample, they just asked the question straight. For another third, they first asked 'How often do you go to the synagogue?' The other third they first asked 'How often do you pray?' In other words, they surreptitiously planted in the minds of these people the idea either of attending religious service or going to church.

It's a simple thing, but it had a dramatic effect. Priming people with thoughts of prayer reduced their support for suicide attacks. But priming people with thoughts of going to a religious service increased it. There was a four-fold difference between the two conditions.

Do these results also hold for people living outside the Palestinian conflict zone? It seems they do. Ginges & co then looked at data from an international survey on religion. This didn't ask about support for suicide bombers, but it did ask about the feelings towards 'in-group' ("I would be willing to die for my God") and out-group ("I blame people of other religious faiths for much of the trouble in this world"). They called this 'parochial altruism' - the idea of sacrificing yourself for your neighbours.

Across the six countries they surveyed (including Hindus, Christians and Muslims and Jews), frequent attendance at religious services doubled the support for parochial altruism. Although prayer increased parochial altruism in some countries, overall it had no effect.

What these results are saying is that going to Church, or to a Mosque, Synagogue or Temple creates a 'coalitional commitment'. On the face of it, this isn't too surprising. After all, we're always being told that religion acts to 'strengthen the community'. And it does. Religious attendance seems to help people to live longer, and to reduce the risk of suicide - presumably by helping them feel part of a special group.

But here we're seeing the nasty flip-side to group cohesion. Religion also helps people to do all the nasty things that they do as groups. If you want to stop society splitting into 'us' and 'them', then the first thing you need to do is to reduce - not increase - support for religious groups. Jeremy Ginges, Ian Hansen, Ara Norenzayan (2009). Religion and Support for Suicide Attacks Psychological Science, 20 (2), 224-230 DOI: 10.1111/j.1467-9280.2009.02270.x


  1. I just wanted to say...this is quickly becoming one of my favorite blogs. Useful and interesting research presented very clearly. Keep up the good work!

  2. Fascinating research.
    I find it interesting that it applies across various countries and religions.
    My first thought was that the specific messages being presented in Israeli mosques and synagogues might be the significant factor but the research seems to imply that the mere fact of attendance at services that has more influence.

  3. It makes you think: Is it the meeting that causes people to egg one another on? Comparisons of board rooms and large corporations behaving badly towards society and the public come to mid, especially the banks recently.

    Has there been any work on the types of decisions made in concert as opposed to solo? Could they be more "extreme" if taken collectively? Or is vice versa true?

  4. Andy, no doubt there's some connection to group think. Talking of which, here's an interesting recent study:

    ONE OF THE most famous phenomena in social psychology is groupthink, the tendency of a group to converge on a consensus without much critical evaluation, even if the consensus is wrong. Various remedies have been proposed over the years, but some management researchers are presenting an interesting new angle on it. They invited a couple hundred members of fraternities and sororities to participate in a problem-solving experiment. The students were given 20 minutes to read a murder mystery and deduce the most likely perpetrator out of three suspects. Individually, only 44 percent of the students got it right, which is slightly better than chance. The students were then sorted into groups of three, all from the same fraternity or sorority, and were given 20 minutes to come to agreement on the most likely suspect. After a few minutes, a fourth member was added to the group - sometimes from the same fraternity or sorority, sometimes from a different one. If the new member came from a different fraternity or sorority, the group performed objectively better then the totally homogeneous groups (75 percent vs. 54 percent correct), and members with incorrect guesses were much more likely to change their minds. Nevertheless, the homogeneous groups perceived themselves as having more confidence, consensus, and effective interaction.


Markup Key:
- <b>bold</b> = bold
- <i>italic</i> = italic
- <a href="">FoS</a> = FoS