The Revival of Experiments on Prayer

Keith Stewart Thomson. American Scientist. Volume 84, Issue 6. Nov/Dec 1996.

The newspapers recently carried a story about a proposal to test, through scientific experiment, the power of prayer. In the careful language of the reporter, the investigator intends to “test what believers have long taken for granted: Praying for sick people may help them get better.” May is the critical word here, scrupulously implying that the opposite could also be true.

Science, of course, is the essence of materialism, of fact and logic. In many respects, science and religion have always seemed opposed, yet how else would one test anything except scientifically, and what better test could there be of the power of prayer than to show its direct material effects-or their absence? On the other hand, how does one set up a strict experiment to test the ineffable and untestable?

In the present case there seem to be all the elements of a scientific experiment. The experimental subjects, pre-term babies (born at 33 weeks or earlier) in three different hospitals will be assigned to three groups of 50: One group will be actively prayed for, one will be ignored, and for the other the prayer will be that “God’s will be done.” (Unstated but implicit is that we are dealing here with the God of Christians.)

Those praying will not know for whom they are praying; they will just get a baby number. Prayers may even be far away in another state. The parents will not know to which group their babies belong.

After a suitable time frame, the health histories of the babies will be compared-growth rates, frequency of serious illness and so on. The numbers will then be analyzed statistically to see whether there are any real differences among the groups. It all sounds very logical and scientific. But even with the best experimental design, I suspect any such experiment is doomed to end in controversy because of a number of problems in procedure and logic.

This is by no means the first time that experiments have been proposed to test the efficacy of prayer. In 1872 in England, an anonymous author (later identified as probably the surgeon Henry Thompson) proposed a grand experiment in this direction. He suggested that “one single ward or hospital” be chosen for prayer over a period of three to five years, and that the progress of patients in areas where the “mortality rates are best known” be compared with “past rates, and also with that of other leading hospitals, similarly well managed, during the same time.” Those praying would be “the whole body of the faithful,” that is, praying would be managed by the Church of England.

Thompson’s experiment was never carried out, but in fact there have been plenty of opportunities to observe whether prayer works. As Frank Turner summarized in an article entitled “Rainfall, Plagues and the Prince of Wales,” the Church of England has a history of ordering special prayers for categories of individual or general need.

Francis Galton (cousin of Charles Darwin) reviewed the results of programmatic praying of this latter sort in 1872 and concluded that it had no effect. Why then would people still want to try to prove the opposite?


The answer is complicated and has as much to do with a reaction by religion to the power of science as it has to do with a pious need to establish the efficacy of prayer itself. In the mid-1800s there was a powerful backlash against the rising power of the scientific method and particularly against the logical materialism of its systems. What Turner describes as the situation in 1872 still applies today: “[Science] . . . had cast into doubt the facts of scripture and the foundations of traditional natural theology … the once friendly and complementary relationship between scientists and clergy gave way to the disputes over Lyell’s geology and Darwin’s evolution.”

The tug-of-war continues between science and religion (organized or otherwise) as to which system ordinary people should attach their loyalties for a better life. It also concerns the ability of individuals to shape their own destinies. Today there is still a strong (and perhaps growing) undercurrent of resistance to the culture of science, especially as it applies to “human nature.” It is not difficult to observe a corresponding increase in interest in angels, parapsychology, ghosts and other nonrational (or at least nonmaterial) subjects. In our local branch of an excellent national bookstore chain, there are more books for sale on “New Age” than on “Nature,” more books on “Angels” than field guides to plants and animals, and more on “Religion” than on “Science.”

Hope and Blood Chemistry

All this is an intensely political contest, and the stakes are huge. To a large extent both scientists and religious leaders hope that the public will accept their authority blindly, but one appeals to the head and the other to the heart. And, for the present at least, that means that it is still a contest fought on different terms.

Matters of the human spirit are not well suited to investigation by the scientific method. Science is to mysticism as a clove of garlic is to the vampire. Scientists tend to concentrate only on those things that they can access through existing material methodology. And since concepts such as hope, love and faith are not (yet) fully reducible to the same terms as infection, inflammation, blood chemistry and molecular genetics, a great deal of human nature is still outside the realm of scientific method.

On the more positive side, however, there is a growing and important interest in the spiritual side of human life. This may play a significant role in medicine, although the medical establishment is still slow to investigate it. Whether “grace” is a state that is conferred by God or achieved simply through human endeavor, there is no doubt that people who are at peace with themselves are happier and healthier, and generally more productive citizens (and, in terms of social support systems, cheaper for the taxpayer). In this realm, investigations of prayer have a legitimate role.

Whether prayer is a subject amenable to any kind of scientific analysis remains a tricky question, however. What seems straightforward is really very complex, but nonetheless fascinating to anyone interested in the experimental method.

A scientific experiment is concerned with material cause and material effect. In the proposed experiments, the effect should be changed health outcomes. The cause appears to be prayer. But is it? When people pray for rain, the rain can only be sent by God. So how can we test God?

Before we can proceed, the nature of “prayer” requires clarification, because the scientist cannot experiment which something that is not properly defined. The “handbook” for Christians and Jews concerning these matters-the Bible-turns out to be less comprehensive on the subject of prayer than one might expect. However, quite clearly prayer involves two (at least) different concepts: meditation/worship and supplication.

A common theme in the Bible, as for most religions, is that the people are urged to spend as much time as possible in prayer. This is prayer of the general meditative type mixed with prayer that is aimed at cementing the general bond between God and his people. For example, “prayer … shall be made for him continually; and daily shall he be praised” (Psalm 72, verse 15). This is both worship and a communication with God-an act of piety designed to earn God’s approval.

Meditative prayer has certain properties, whether or not there is any God to hear the prayer. Evidence is growing that the experience of sharing prayer with friends has a powerful effect on health outcomes. These benefits are susceptible to explanations that do not necessarily involve the direct intervention of God.

But experiments on the efficacy of prayer have always depended on the more difficult, and entirely different, case of supplicative prayer-prayer for a specified outcome such as rain. Such prayer is not the causal agent of change, but simply the intermediate agency-a supplication to God.

Supplicative prayer is more common in the New Testament (which refers to a more personal God); nothing could be more familiar than “Ask and it shall be given you; seek and ye shall find” (Matthew 7:7). However, even this seems to be referring to things spiritual rather than material. When his disciples asked Jesus how to pray, he gave them the Lord’s Prayer, in which the only selfish request is rather basic-for “daily bread” (Matthew 6:9-13). Nothing about health, rain or personal gain.

Jesus is reported by both Mark and Matthew (the one being based on the other) in very clear terms. “Asking” is useless without faith: “Whatsoever ye shall ask in prayer, believing, ye shall receive” (Mark 11:24; Matthew 21:22). The emphasis here (the figtree miracle) is on “believing.” And this is reinforced by James (1:6): “But let him ask in faith, nothing wavering [‘doubting’ in newer translations]. For him that wavereth is like a wave of the sea driven with the wind and tossed. For let not that man think that he shall receive anything from the Lord.”

All this, therefore, is still quite different from the more egregious kind of prayer positing God as a kind of personal genie, exemplified by one Elias who “prayed earnestly that it might not rain, and it rained not on the earth by a space of three years and six months. And he prayed again, and the heaven gave rain….” (james 5:15-18). It seems unlikely that James’s readers would have believed this story any more than we do.

Putting God to the Test

This, then, is the ultimate test of prayer-whether it can successfully call God to cause rain in a drought or relieve sickness, and particularly whether it can produce a result on someone who does not even know that he or she is being prayed for.

However, if it is true that for supplicative prayer God requires the supplicant to have faith that the prayer will be answered, and if a doubter will not have his or her prayers answered, then any true experiment on prayer logically must fail. Science requires an open mind, not faith. An experiment, a test, implies doubt, and it is the curse (or blessing) of the scientist to be the ultimate doubter. Doubt logically invalidates prayer. Equally, faith obviates the need for data, which removes us from the realm of science.

A further logical problem concerns the experimental group that will be praying “that God’s will be done.” On the one hand this is something like King Canute reversing his own famous experiment by saying to the sea, “I command the tide to flow!” The results for this group should not differ from those of the control group with no prayer. But it does raise the question of how the result for the “prayed-for” group could in any case be different from God’s will. A God that is all-wise and all-knowing (to say nothing of all-powerful) presumably has a much better idea about what is the right course of action than a bunch of presumptuous human beings. (Of course, that would certainly accord with the fact that supplicative prayers do not seem to be answered.)

On the other hand, a God that can be entreated or even bargained with is less all-powerful and perhaps unreliable. In other words, the experimental design makes a series of theological assumptions that are themselves untestable.

On the methodological side, there is a question of controls. It is not just that the effective agencyGod-may be totally unpredictable; the intermediate agency-prayer-is not realistically controllable. In the proposed experiment with babies, the parents of the “no-prayer” babies cannot be told not to pray for their children, because that would violate the experimental protocol by informing parents which group their babies were in. In fact, whether it be parents, grandparents or the “general prayers for the sick” in churches on Sunday, all the babies might be prayed for anyway.

So this raises the problem of dosage. How much prayer is enough? Would God not be far more likely to hear one genuine prayer from a mother, over a hundred experimental prayers?

Equally obviously, any experimental design depends on having appropriate experimental samples. Groups of 50 differently premature babies will scarcely suffice. Ideally, for such small numbers the babies should have comparable prenatal histories, postnatal care and exposure to disease and accidents. They should have the same degree of praying-over. In other words, unless huge numbers of subjects are available for randomness, the experiment might better be confined to identical twins, assigned to opposite groups-prayed-for and not-prayed-for.

Perhaps the most fundamental problem is that experiments to test the power of prayer on unwitting subjects at long range, if any effect is to be caused, require God or some other agency to act through non-material means. In short, to work miracles. The whole enterprise is therefore not really about prayer as a practice, but about the existence of God. And while it is perfectly rational to want to test whether God exists and whether miracles happen, no result-negative or positive-will be accepted by those who wish to believe otherwise.

At this point one begins to question whether all such experiments come close to blasphemy. If the health outcomes of the prayed-for subjects turn out to be significantly better than for the others, the experimenter will have set up a situation in which God has, as it were, been made to show his (or her) hand. It is an attempt to maneuver God into an irrefutable scientific “proof” of his (or her) own existence-something that he or she has hitherto required people to accept either on authority or as a matter of faith.

This also revives the interesting question that philosophers have argued for hundreds of years: If material proof of God could be obtained, would faith any longer be necessary? And in that case would not God, in the end, be diminished or destroyed rather than magnified?

For the deity portrayed in the recent literary “biography” of God by Jack Miles, this sort of messing around with God could prove dangerous. In looking through Miles’s index, categories such as “commandments,” “conqueror,” “fear and terror of” and “unknowable” are well represented, but I would feel better if a “sense of humor” were there too. Meanwhile, experiments will proceed and no one will be satisfied, perhaps especially those who remember that, in the words of Paul, “God is not mocked” (Galatians 6: 7).