Does seeing things as if you were God rather than yourself make you less biased towards members of outgroups? Acccording to a new paper in the Proceedings of the National Academy of Sciences by Jeremy Ginges et al. (reference below, free access, and Scott Atran is a co-author), the answer is “yes.” This was the result of a psychology experiment conducted on 555 Palestinian children between the ages of 12 and 18.

The study was motivated by wanting to know whether taking God’s viewpoint made the subjects more moral, at least in terms of valuing people’s lives equally:

For religions such as Islam that believe in a moralizing God capable of punishing immoral behavior, thinking about God should invoke religious norms. If, on the whole, religious belief devalues the lives of nonbelievers, placing them outside the scope of moral concern, then people should believe that God prefers parochial choices that imply a decrease in the relative value of the lives of nonbelievers. However, if God is seen as promoting universal moral laws, and if those laws conflict with intergroup bias, participants should believe that God would prefer choices that more equally value human life regardless of religious identity.

Now given that the Qur’an repeatedly devalues the lives of nonbelievers, to the extent of mandating their deaths, you might wonder why belief in Allah (whose words are the Qur’an) would lead you to become more tolerant of unbelievers when taking His viewpoint rather than your own. But never mind; we can ponder the meaning of the results after we see them.

The 555 Palestinian children were equally divided among males and females, with 64% coming from the West Bank and the rest from Gaza. Ginges et al. claim, reasonably, that their experiment was best conducted in areas of inter-group violence, which may promote more acute thinking about moral duties.

The children were presented with a modified version of the classic “trolley problem,” a moral dilemma made famous by Phillippa Foot and Judith Jarvis Thompson. You probably know it: a person is facing a situation in which five people are on the track of a runaway trolley and sure to be killed. But if you throw a switch, you can divert the trolley onto another track where there’s only one person. Do you throw the switch, saving five lives but sacrificing one? Most people answer “yes”: it’s the right thing to do.

An alternative version requires you not to flip a switch, but to throw a fat person off a bridge onto the tracks to stop the trolley, again saving five lives but taking more direct action to sacrifice one—via physically throwing someone to his death. (A “person of size” is specified because you yourself are supposed to be too light to stop the trolley.) As I recall, in this case most people say it’s immoral to throw that person to his death, even though the result is the same as with throwing the switch. The difference is that people think the “throwing the fat man” method is more immoral because you’re directly harming a person by laying hands on him. Regardless of what you think, the difference in answers tells us something about our innate moral reactions.

Since the Palestinian children weren’t familiar with trolleys, the researchers used a runaway truck instead. The problem posed was this: do you jump off a bridge yourself to stop the truck and save five children? And does it matter whether the five children are Palestinian or Jewish Israelis?

Another version of the dilemma was posed (these were divided evenly among the children): facing the same truck, do you throw a fat Palestinian in front of it to stop it? If so, do you prefer to save the Israeli or Palestinian children?

As I said, these two dilemmas were divided among the children. Surprisingly, the results didn’t differ between the “jump” and “push” questions, so the data were combined. But in both cases, all children had to answer their question for both Palestinian victims and for Israeli victims.

Then each group of children had to answer the same question again, but this time from the viewpoint of Allah. Here’s how the questions were posed (they were posed by two Palestinian investigators, in case you were wondering):

One more note: the authors add that “The majority of our children were very religious and prayed regularly (>80%).”

I’ll be brief with the results, encapsulated in the table below. Remember that the results of the “throw” and “jump” questions were lumped bcause they didn’t differ among groups.

There are two rows, with “self” representing how children answered when they were to make the judgment themselves, and “Allah” representing which action they thought Allah would approve of. “Ingroup bias” repesents those children who said that either they or God would favor the Palestinian children over the Israeli ones, “no bias” are respondents who showed no preference between potential Palestinian or Israeli victims in either saving or killing the children, and “outgroup bias” are those who approved of saving Israeli but not Palestinian children:

As you see, thinking about the dilemma from Allah’s prespective led the children to reduce their ingroup bias by 28% (12/42), making them more neutral about the value of Israeli versus Palestinian lives. The proportion favoring the lives of Israeli children, which as expected was low (3%), didn’t change. The authors say this:

These results reveal that participants believed that they had preferences different from those of God when it came to answering certain moral dilemmas. Rather than encouraging divisive tribalism, participants believed that God had relatively stronger preferences than they did to treat the value of human lives equally, regardless of religious identity.

There are other analyses, but the table above gives the important result. One other finding, though: they asked the children another question and correlated its answer with the answers above:

To demonstrate the relevance of our ingroup bias measure for understanding intergroup violence, we regressed “yes” answers to a question asking participants whether they thought it was their “duty as Muslims” to kill nonbelievers, on an overall measure of ingroup bias for evaluation of lives (average of self and God’s perspective). We created dichotomous measures for ingroup bias giving a score of 1 if participants showed ingroup bias and a score of 0 if they did not. Ingroup bias was related to approval of intergroup violence; those who showed bias were more likely to believe it their duty to kill nonbelievers (B = 0.45, SE = 0.18, z = 2.52, P= 0.01).

If that’s TL/DR, children who thought it was their duty to kill nonbelievers were more likely to show ingroup bias whether they took their own point of view or that of Allah. Curiously, the authors don’t say how many children actually answered “yes” or “no” to the question about killing nonbelievers.

Ginges et al. conclude from their study that there’s nothing special about religion that “invariably” (note the weasel word) promotes violence against outgroups. To avoid distorting this, I’ll give their conclusion directly from the paper (my emphasis):

Humans will fight, kill, and die for a variety of abstract beliefs or entities, including national rights and ideological doctrines of many types. Together with intriguing prior work showing how priming God can increase cooperation with outgroups, or how costly signaling of religious belief can increase trust across religious boundaries, our findings cast doubt on the notion that there is something special about religious faith, including Islamic belief, that invariably favors promotion of violent intergroup conflict.

This sounds a bit like extrapolating their findings beyond the data, even if they do consider religious faith as an “ideological doctrine”.  And, given that most of the children were pious Muslims to begin with, how do we interpret the change in attitudes when they take God’s point of view? Do religous people normally take God’s point of view?

And even when they do interpret Allah’s wishes, the religious bias doesn’t go away. If that were the case, then nearly 100% of the children would fall into the “no bias” column in the second row. What we see is that 66% are in that group (up from 55%, not a huge effect), but that the preference for Palestinian over Israelis lives is still in the ratio of ten to one.

I’d like to see further questions alone these lines, although this study was part of a larger survey. What would the children say if asked whether they (versus Allah) would approve the killing of infidels, adulterers, or apostates? Would they become less tolerant of violence when taking Allah’s viewpoint?

In the end, I’m not sure what to make of these results, except for the data themselves: if you take Allah’s viewpoint, you’re less approving of outgroup violence. Will that change the children’s minds if they think about it?

And how do we translate these results into action? Is it even practical to ask kids to think not on their own, but as if they were God? As for the conclusions just above in bold, I don’t know what to make of the authors’ statement that there’s nothing “special” about religion or Islam that promotes violence. Maybe not, if you construe “faith” as a brand of divisive and unthinking ideology. But unless you’re of the Armstrongian or Greenwaldian stripe, religious faith does inspire violence, and—or so I think—Islam does it more often than other religions.

Ginges, J., H. Sheikh, S. Atran, and N. Argo. 2016. Thinking from God’s perspective decreases biased valuation of the life of a nonbeliever. Proceedings of the National Academy of Sciences 113:316-319.