Prayer

What it is

Address to the divine — and the word covers activities so different that evaluating "prayer" as one thing repeats the mistake made with meditation.

The main kinds:

• Petitionary — asking for something for oneself
• Intercessory — asking on behalf of another
• Contemplative — wordless attention to God, closer to meditation than to asking
• Liturgical — set forms, said communally, at appointed times
• Thanksgiving, praise and confession — none of which requests anything

Only the first two make claims that a trial could test. The others are doing something else entirely, and criticising prayer by testing intercession is like assessing music by measuring whether it moves furniture.

Where it came from

Universal. Every religion this site covers prays in some form, and the forms are old: fixed hours, prostration, repeated formulas, silent interior address, communal recitation.

Two features recur widely. Repetition — the rosary, the japa mala, Orthodox prayer ropes, Islamic tasbīḥ — which shades into the practices covered under dhikr and the Jesus Prayer. And fixed times, which impose a shape on a day regardless of inclination.

How it is done

Spoken aloud or silently, alone or communally, spontaneously or by set form, in prescribed postures, at prescribed hours, with or without beads.

What we can and cannot say

We can report the intercessory prayer trials precisely, because they were done properly.

The largest is the STEP trial — Study of the Therapeutic Effects of Intercessory Prayer — led by Herbert Benson and published in the American Heart Journal in 2006. It enrolled 1,802 coronary artery bypass patients across six hospitals, randomised into three groups:

1. Received prayer, told they might receive it
2. Did not receive prayer, told they might receive it
3. Received prayer, told certainly that they would

Congregations were given patients' first names and prayed for an uncomplicated recovery.

The results were unambiguous and contained a surprise. Prayer made no difference to complication rates between groups 1 and 2. And group 3 — the patients who knew they were being prayed for — had a higher rate of complications than either other group.

The most-discussed explanation is performance anxiety, or what the authors' commentators called the "am I so sick they had to call in the prayer team?" effect. Whatever the cause, the one statistically notable finding ran the wrong way.

A Cochrane review of intercessory prayer trials reached an inconclusive verdict and suggested that further trials of this kind were not a good use of resources.

And the theological objection deserves stating fairly, because it is not a dodge. A great many theologians regarded these trials as misconceived from the start. The design treats God as a mechanism that responds to inputs at a measurable rate — precisely the transactional model that sophisticated theology in most traditions rejects. On the classical view, prayer is participation in a relationship, not the operation of a lever, and a God who could be shown to increase recovery rates by a reliable percentage would not be the God being prayed to.

This site takes that objection seriously. It also notes what follows from it: a claim removed from testability cannot then be offered as evidence when someone recovers.

We can say prayer does other things that are measurable. Religious participation correlates with better health and longevity in large cohort studies — though most of the effect appears attributable to social support, behavioural norms and community rather than to prayer specifically, and disentangling them is difficult. Personal prayer functions for many people as a structured practice of reflection, gratitude and grief, and those have their own literatures.

We cannot say anything about whether prayer is heard. What can be said is that its effects on the person praying are real and various, that its effects on distant strangers have been looked for carefully and not found, and that the most interesting result in the whole literature is that telling someone they are being prayed for is not obviously good for them.

Further reading

• Benson et al., "Study of the Therapeutic Effects of Intercessory Prayer" (American Heart Journal, 2006).
• Roberts et al., "Intercessory prayer for the alleviation of ill health" (Cochrane review).
• Any serious theological response to the STEP trial — the objections are more interesting than the trial.