"Religion," novelist Mary McCarthy wrote, "is only good for good people." Weigh the violence of the Inquisition against the humanity of Martin Luther King or homicidal fanatics against Oxfam, and you have to suspect that religion supplies a context for justifying or motivating moral choices rather than a reason for them.

Into this bitterly contested arena comes a new paper by psychologists Ilkka Pyysiäinen of the University of Helsinki and Marc Hauser of Harvard University in Cambridge, Massachusetts1. They point out that individuals presented with unfamiliar moral dilemmas show no difference in their responses if they have a religious background or not.

The study draws on tests of moral judgements using versions of the web-based Moral Sense Test that Hauser and others have developed at Harvard2,3,4,5,6. These tests present dilemmas ranging from how to handle freeloaders at 'bring a dish' dinner parties to the justification of killing someone to save others. Few, if any, of the answers can be looked up in holy books.

Good by nature

Thousands of people — varying widely in social background, age, education, religious affiliation and ethnicity — have taken the tests. Pyysiäinen and Hauser say the results (mainly still in the publication pipeline) indicate that "moral intuitions operate independently of religious background", although religion may influence responses in a few highly specific cases.

Atheists behave as well as believers in many moral tests. Credit: Michelangelo Buonarroti/ Tomisti

This finding may speak to the origins of religion. Some researchers have suggested it is an adaptation that promotes cooperation between unrelated individuals7,8 — by, for example, discouraging cheating with the notion that 'God is watching'. Others say that religious behaviour is not specifically evolutionarily selected for in human evolution, but arises as a by-product of other cognitive functions and capacities9,10: religion may, for instance, have appropriated underlying psychological reasons for a belief in souls and an afterlife.

Because religion has little influence on moral judgements, say Pyysiäinen and Hauser, the latter hypothesis seems more likely. They argue that human populations evolved moral ideas about behavioural norms — which themselves promoted group cooperation — before they became encoded in religious systems. The researchers suggest that we may possess an innate 'moral grammar' that guides these ideas.

The paper plays to a wider issue than this point of largely anthropological interest, for it challenges the assertion commonly made in defence of religion: that it inculcates a moral awareness11. If we follow the authors' line of thinking, religious people are no more likely to be moral than atheists.

Pyysiäinen and Hauser do not wholly deny that religion is adaptive. They think that natural selection may have fine-tuned it, from an existing array of moral-determining cognitive functions, to optimize its benefits for cooperation. There is some evidence that religion promotes in-group altruism and self-sacrifice beyond that displayed by non-believers12,13.

Core beliefs

The authors' paper may annoy both religious and atheistic zealots. By taking it as a given that religion is an evolved social behaviour rather than a matter of divine revelation, it tacitly adopts an atheistic framework. Yet at the same time it assumes that religiosity is a fundamental aspect of human psychology, thereby undermining those who see it as culturally imposed folly that can be erased with a cold shower of rationality.

It's debatable, however, whether these moral tests are probing religion or culture as a moral-forming agency, because non-believers in a predominantly religious culture are likely to acquire the moral predispositions of the majority. Western culture, say, has long been shaped by Christian morality.

All the same, the tests show that neither culture nor religion matter very much: other factors — presumed to be inherited — dictate our judgements.

Certainly, religious moral doctrine sometimes displays such inconsistency that you have to suspect it is being shaped by unspoken prior judgements — rather than religious tenets as such. Take, for example, the Catholic church's early opposition to in vitro fertilization, which sat alongside an otherwise fierce prohibition of any hindrance to procreation. And most religions have the same set of core moral principles about lying, theft and murder, all with evident adaptive benefits to a group, beyond which the details (Christian original sin, say) are a question of historical contingency.

But to uncover religion's roots, is morality necessarily the best place to look? It seems hard to credit the idea that the immense cultural investment in religion was made merely to strengthen and fine-tune existing neural circuits related to morality. Some people place more emphasis on the adaptive rationale for religious symbols and mystical beliefs, rather than morals10.

Yet attempting to explain the origins of such a rich cultural phenomenon as religion is doomed to some extent to be a thankless task. For to 'explain' Chartres Cathedral or Bach's Mass in B Minor in terms of non-kin cooperation is obviously to have explained nothing.