You are here

1: The Prayer Experiment

Does prayer work?

You are suffering from heart disease-in need of coronary artery bypass surgery. At the hospital someone asks whether you are prepared to take part in an experiment. Not, I hasten to say an experiment involving some novel surgical technique; the operation will be perfectly standard. No, it is an experiment to test whether prayers for the sick are effective.

On agreeing to participate, your name is to be randomly assigned to one of two groups. Patients in one group will be prayed for by special teams of intercessors drawn from a variety of religious denominations, each team being given a particular batch of patients. Patients in the second group will not be prayed for. The second group is identical in all respects to the first, with the sole exception of its not being the subject of prayer. There will, of course, be the inevitable statistical differences between the groups, but these should be small, given that each group consists of 600 patients drawn from the three major US hospitals participating in the experiment. The second group is the control group.

Over a period of two to three years, the patientsଁ case histories will be followed up to see if there are any differences in recovery rates between the two groups. Indicators will include measurements of the physical functioning of the heart, the frequency of death from all causes, length of stay in hospital, and whether patients discharged from hospital return to their homes or enter a nursing home.

As a participating patient you will not know to which of the two groups you have been assigned. You might be prayed for, or you might not. That is all you know. Likewise, the medical staff tending you will have no access to this information. Those doing the praying will know only the Christian names of the people assigned to them, together with some details of their condition - not sufficient to reveal the identity of the individual. Only at the very end of the experiment will all the data be collated. The project has been designed as a rigorously controlled scientific experiment.

I first came to learn of this experiment through being a trustee of the John Templeton Foundation, an American charitable organization devoted to supporting projects promoting progress in religion. It was our foundation that agreed to fund this experiment. When I recently spoke of the project in a radio broadcast, I was inundated with queries from newspaper reporters and radio and TV interviewers; the telephone did not stop ringing for a week. Everyone seemed totally amazed-and intrigued-at the thought that one might apply scientific principles to investigate a religious question.

But why? Several other fields of academic activity have tried-admittedly, with varying degrees of success-to turn themselves into sciences. We have seen, for example, the rise of the social sciences; we have heard Sigmund Freud and Carl Jung claiming on behalf of psychology that it too was a science. So why not theology?

One objection stems from the idea that theology is fundamentally rooted in Scripture. The starting-point is that one accepts, in faith, that the Bible, or some other set of sacred writings, is the inerrant Word of God. This set of writings, interpreted in a literal manner or in some more liberal way, is the ultimate authority on all matters relating to God.

If that is the stance adopted, then it is clearly unlike anything that holds in science. Science is fundamentally based in experimentation and observation. It is an approach where one's theories are in a constant state of flux as they adapt themselves to new empirical findings. No matter how cherished one's hypothesis might be, it must always yield in the face of contrary evidence. The ultimate authority lies with the data. The idea that the truth was set in stone at some earlier time, and is not to be questioned, has no place in science.

But that is just one approach to theology. There is another-one that resonates much more closely with the scientific approach. One of the tasks of this book is to explore to what extent this second approach to theology can be regarded as a science.

I suppose most people's idea of the way scientists go about their work is of an investigation conducted under carefully controlled laboratory conditions. A situation is set up in which, as far as possible, the particular effect under investigation is isolated, and all other extraneous background effects are reduced to a minimum. That is indeed the aim in many scientific studies. As regards religious investigations, the best attempt to apply such a methodology is to be found in the experiment I was talking about just now-the so-called prayer experiment. That will be our starting-point. So let us take a closer look at it.

The prayer experiment

Belief in the power of prayer is widespread. One hears the claim, for example, that the collapse of atheistic communism was an answer to prayer. It had long been held that, because of the totalitarian nature of communism-the lack of free elections-once a country went communist there could be no going back. Religion in those countries would become a thing of the past. And yet it was not to be. Against all the odds, communist regimes have toppled, and the churches in those countries, suppressed for decades, are now coming back into their own once more.

Or take the case of South Africa. It was an almost universally held belief that apartheid in South Africa could end only in a bloodbath. In the event, the transition to majority rule occurred relatively peacefully. This is also claimed to be an answer to prayer.

Though such examples might strike believers as highly persuasive, they are not, of course, conclusive. We are unable to rerun history-once with the benefit of prayer, and once without-to see if it makes a difference. No, to try and see whether prayer ‘works’, one has to turn away from the big one-off world events, and concentrate on situations that are repeatable - frequently occurring situations where we have a chance of varying the conditions.

Prayers for the sick appear to offer just such an opportunity. If miraculous healings happen, then this would seem on the face of it to be evidence for God - God directly intervening and having an effect in the world. There is a widespread acceptance of the efficacy of intercessory prayer. This might be evident in the spectacular, highly charged, emotional rallies held by the evangelist Maurice Cerullo, or in the more sober practice of making pilgrimages to holy places such as Lourdes in France or Walsingham in Britain. More commonly, faith in miracles manifests itself in the intercessions for the sick offered up in every church on a Sunday, or in the daily private devotions of millions when praying for the recovery of sick loved ones. But do such prayers work? That is what the prayer experiment hopes to discover.

The project's chief investigator is Dr Herbert Benson of the Mind/Body Medical Institute, New England Deaconess Hospital, Boston, Massachusetts, USA. This is not the first time an experiment of this sort has been carried out. Results published in 1988 of an intercessory study performed by R. C. Bird appeared to show a significant difference between the two groups of patients. Those who received intercessory prayer appeared to have a less complicated hospital course and required fewer medications and procedures than those belonging to the control group who did not receive prayer. However, the results have not been widely accepted, in part because of concerns over the details of the methodology involved.

The current project is a considerable improvement on the original in a number of respects. The sample sizes have been increased from the 200 characteristic of the first study to the 600 involved in the present experiment. Secondly, whereas the Bird study included anyone admitted to the coronary care unit, the patients involved this time are more homogeneous in that they are all undergoing the same surgery (thus facilitating tighter medical definitions of evaluation points or end-points of treatment). The follow-up period has been extended from one year to two/three years. Three sites are involved instead of one. More sophisticated statistical techniques will be employed to analyse the data.

In addition, the Benson study includes a third group of 600 patients. Like those in the first group, these are prayed for, but unlike the first, they are informed that they are to be the subject of prayer. The intention here is to examine whether there is any additional benefit to be gained (of a psychosomatic or placebo nature) from knowing that one is the subject of prayer. This part of the study will test hypotheses to do with ‘patient expectation’.

Possible outcomes and interpretations

And what of the results? It is too early to say. At the time of writing, the experiment has only just got under way. It will be two years before there are any indications as to how it is going. So, you might be wondering, why bring up the subject now? Are we not being premature? The reason is that, long before it was agreed to go ahead with the experiment, much thought had been devoted to considering the possible outcomes, and what significance might be attached to them. Even without knowing the actual outcome, it is highly instructive to think through the various options.

As regards the first two groups - those unsure whether or not they are being prayed for - should the project eventually yield a statistically significant positive correlation between prayer and good recovery from the operation, that would of course be fascinating. Doubtless such a discovery would trigger a whole series of follow-up experiments. In the first place one would want to establish, beyond all reasonable doubt, that the positive correlation was no mere statistical freak. Different prayer techniques would be tried to see whether some were more effective than others. One would want to investigate whether other medical conditions yielded to this form of treatment - cancer, perhaps. Clearly an important field of study would open up.

However, it must be noted that none of this would amount to proof of God's existence; persuasive evidence in favour, maybe - but not out-and-out proof. Both the investigators themselves and the funding agency make it clear from the outset that the project is not to be thought of as an attempt to prove God's existence. An alternative explanation of a positive correlation might involve, for example, some form of direct transference of thought between the mind of the person praying and that of the patient, presumably of a telepathic nature. There would accordingly be no need to invoke God as an ‘intermediary’ in the process. The beneficial effect would pass directly from intercessor to patient.

As for the other possible result - no measurable difference between the two groups - what might be the implication of that?

Doubtless many will jump to the ‘obvious’ conclusion: intercessory prayer does not work - perhaps because there is no God. But again one must be careful. In the same way as a positive result does not necessarily vindicate belief in God, so a negative result does not have to be damaging to belief. There are alternative explanations of a null result:

For instance, when it is said that one of the groups will not be prayed for, that simply means there will be no special prayer team at work on their behalf. That of course will not stop the patients praying for themselves, nor their loved ones and friends from praying. The investigators refer to this (somewhat tongue-in-cheek) as the ‘unwanted background noise’. What the experiment is trying to do is to measure whether there is any additional benefit coming from the prayers of the special team. It could well be that the efforts of these strangers will be swamped by the heartfelt prayers of those directly involved with the patients.

There is another concern: when scientists investigate the physical world, provided they ask the right questions and adopt good, sound scientific methodology, nature has no alternative but to yield up its secrets. But applying that methodology to God (or to anyone else with a will of their own) is not a guarantee of success. God might simply decide not to cooperate. It could be argued that a loving God might indeed be reluctant to penalize patients merely because some strangers deliberately choose not to include them in their prayers.

Not only that, God might well appreciate better than we ourselves some of the drawbacks of allowing us to probe too deeply into his manner of working. By the very nature of the clear-cut way in which the experiment has been designed, the result will be a quantitative one. That means it becomes possible to put a price on whatever benefit might be gained from this type of prayer exercise. Do we really want some treasury official to use a positive result as an excuse to close down hospital wards - because vicars can be hired to pray more cheaply? While we are doubtless all in favour of a cost - effective National Health Service, it seems a far cry from what many would regard as the true nature of prayer: a natural expression of a loving, trusting, personal relationship with God. For this reason, it would not altogether be surprising if God decides to frustrate the best efforts of the investigators - for their own good. We would do well to recall that it does say in the Bible: ‘Thou shalt not put the Lord thy God to the test.’ So, we cant say we haven't been warned!

Important and intriguing though this experiment might be, it is necessary to keep it in perspective. The experiment is concerned solely with intercessory prayers; there are many other aspects of prayer. Moreover, it deals only with prayers offered up on behalf of strangers. Like many others, I reckon the central core of intercessory prayer has more to do with the agonizing, involved prayers of loved ones and intimate friends - the so-called ‘noise’ - than with those of distant strangers.

Nevertheless the Benson experiment has aroused enormous interest, it has been beautifully designed, and it deserves to be done. I for one await with great interest both the outcome itself, and the heated discussions that are bound to ensue as to what interpretation should be put on the result - whatever that turns out to be.

Limitations of solitary investigations

In view of the ambiguities that will inevitably attend the result of this experiment, you may be wondering why I have chosen it as the starting-point for our discussion of the evidence for God. The reason is simple. It is to bring home that even when we are able to set up conditions of our own choosing, under thoroughly scientific constraints, it becomes surprisingly difficult to arrive at any firm, incontrovertible conclusion as to whether or not a supernatural intervention has taken place. If we cannot get knock-down proof when we are controlling everything, it is surely unreasonable to expect any single, decisive proof of God's existence, and of his influence in the world. Those who make such a demand are likely to be disappointed.

So does that mean that the whole idea of trying to apply scientific thinking to the study of God is ill-conceived? Is it doomed to failure from the outset? No. It is important to recognize that the kind of difficulties encountered over the interpretation of the outcome of the prayer experiment are nothing unusual; they are to be found all the time - even in hard-core physical science.

Take, for example, my own field of research: high-energy nuclear physics. This is the study of the behaviour of sub-atomic particles. These particles are so small they cannot be seen directly (under a microscope, say). They are to all intents and purposes invisible - much like an invisible God. Instead, one has to observe their effects, and from these draw inferences. The usual effect is that they leave a trail or track behind them marking where they have been. (The track might be in the form of a string of tiny bubbles in a transparent liquid, or a series of small sparks in a gas subject to a high electric voltage.) As the subatomic particles collide with each other, or as they spontaneously disintegrate, so they leave behind characteristic patterns of tracks marking out the paths followed by the various particles involved.

The trouble is that for any given pattern of tracks, there might be a number of competing interpretations. If all one has is a single event, there might be very little that can be deduced from it. Many examples might have to be collected, perhaps using a variety of detection techniques; statistical analyses have to be carried out as to the likelihood of the various rival explanations. Only after a period of time might a consensus emerge as to which is the more likely interpretation. Throughout this process there will not be any well-defined point at which the interpretation is proved correct. Rather, the evidence progressively accumulates; the odds become stronger in favour of one particular explanation. Eventually, the controversies die down, and the scientific community as a whole finds that it is no longer seriously questioning the conclusion.

As another example, we can look to cosmology. This will figure prominently in our discussions later, in Chapter 9. There we shall be examining the evidence for the Big Bang origins of the Universe. According to this theory, the Universe began with a great explosion. Why do we think that? For a start, we find that the Universe is expanding; it is still expanding in the aftermath of that explosion. At least, that is the interpretation put on it. But there is another possibility. According to a rival theory, new matter is continually being created throughout space. As fast as matter moves away, the gaps left behind are filled by the newly created material. Thus, the overall picture remains essentially unchanged over time: in particular, there was no explosive beginning. This ‘Steady State’ theory used to be a serious rival to the ‘Big Bang’ theory. And that is how the situation would have remained, had not other indications come along.

As we shall be seeing later, the Big Bang theory holds that the initial conditions in the Universe would have been very hot; the explosion would have been accompanied by a fireball. It is argued that the cooled-down remnants of that fireball ought still to be about in the Universe today. Indeed, this radiation has now been discovered. At least, it has all the hallmarks expected of it. But of course there are many sources of radiation in the Universe. It could well be argued that, just as the expansion of the Universe on its own was not proof of the Big Bang hypothesis, neither is this form of radiation, on its own, clinching proof. Rather, it is to be seen as helping to strengthen the overall case.

The expected conditions of the Big Bang also allow one to calculate what kind of atomic particles were likely to emerge from it. The calculated abundances of the different kinds of atom are found to be in good agreement with those of the atoms that make up the gases from which the stars formed. This constitutes yet further confirmatory evidence.

Finally, we note that it takes time for light to reach us from the far depths of space. Examining distant objects through a telescope is like looking back in time. We are able to see directly how the Universe was long ago - and it looked different then to what it does now. The density of matter throughout space appears greater then than it does now - in agreement with the Big Bang theory, and contrary to what would be expected on the basis of the Steady State theory. But in saying that, it has to be recognized that there are difficulties involved in trying to estimate distances to far-off astronomical objects. Systematic errors could creep in, and if present, would distort one's estimates of density. For this reason, the evidence, once again, is not by itself conclusive.

What one discovers from a discussion like this is that acceptance of the Big Bang theory does not rest on any single decisive piece of evidence. Instead, the case has to be based on the examination of a range of indicators, all of which point to the same conclusion. The Big Bang hypothesis is an economical and elegant way of accounting for a wide variety of disparate phenomena. The evidence is cumulative; it is persuasive rather than clinching. And different scientists need different degrees of persuasion in order to be won over.

One lesson we can draw from this is that, when trying to establish the reality of God, rather than looking for one decisive experiment - say, the prayer experiment - we perhaps ought to be broadening our horizons. We need to examine a wide variety of indications to see whether, taken together, they add up to a compelling case.

When conditions are not under our control

A further reason one might have for believing it inappropriate to associate scientific methodology with studies of God is that, apart from the prayer experiment, it is hard to think of situations that would lend themselves to the establishment of controlled investigations into God.

This, in fact, is not a problem peculiar to theology. An inability to control the circumstances of an investigation is quite common in science too. Take, for example, cosmology - the study of the Universe on the largest scale. We have just been discussing the scientific evidence for the Big Bang. But, of course, there was never any question of us controlling the circumstances of the Big Bang. The cosmos cannot be contained in a laboratory. In fact, the whole subject of cosmology and astronomy is one where we control virtually nothing. We are reduced to observing, in a purely passive manner, whatever radiation signals happen to come our way.

And yet no one would deny that cosmology and astronomy are ‘sciences’. They gain that status from the systematic and scrupulous way the data are gathered, and the manner in which the theories are always open to modification in the light of those empirical findings.

The same holds in certain branches of the biological and earth sciences. In Chapter 10 we shall be examining Darwin's Theory of Evolution by Natural Selection. We shall see how we humans, in common with the other animals, have descended from more primitive ancestors - stretching all the way back to the inanimate chemicals to be found on the surface of Earth soon after it formed some 4,600 million years ago. This is the generally accepted scenario for human origins.

But how has this assessment been made? The evolution of humans is not a repeatable process. Indeed, we shall later be pointing out that if evolution has taken place on other planets in the cosmos, giving rise to intelligent life-forms, these are most unlikely to end up looking like us humans. And yet evolutionary biology is most definitely a ‘science’. The reasons are the same: the systematic collection of a wide range of different types of data (the fossil record, anatomical comparisons between species, genetic comparisons, etc.), together with a willingness to allow the weight of that empirical data to be the final arbiter on the success or otherwise of the theoretical hypotheses. Nor is it necessary for that data to be so overwhelming as to convince everyone. There are many professional biologists, let alone lay persons, still unhappy about various aspects of the conventionally accepted evolutionary theory.

Finally, we might mention the earth sciences. These have been transformed in recent times by the theory of plate tectonics - the manner in which various parts of the Earth's crust move about the surface of the globe causing earthquakes, and the building of mountains. Once again, these are processes that are not under our control.

Thus, the ability to control the circumstances under which an experiment is conducted, together with the possibility of repeating the experiment under the same or different conditions, though desirable, is not a necessary prerequisite for that field of study to be regarded as scientific. For that reason we speak of the social sciences - despite the strong limitations that exist on the extent to which it is acceptable to manipulate social groupings. We might also speak of psychology as a science, even though there are strict ethical limits on the degree to which a therapist might seek to ‘experiment’ on his/her clients.

Opening out the prayer experiment

The same kind of situation confronts us over the study of God. God is not some object we can control. We cannot expect God to submit to repeated experimentation under conditions of our choosing. There will be no single, definitive proof of his existence and of his manner of working. Instead, we have to examine, as objectively as we can, a whole range of experience - a multitude of diverse indications - whatever happens to be on offer. We have to ask: given the totality of the data presented to us by our experience of life, does it make better sense in the light of the God hypothesis? Does it all hang together better if we assume there is a God? If so, what kind of God are we dealing with?

Thus, the line of investigation that began with a narrow experiment into the effectiveness of intercessory prayer is opened out. The prayer experiment becomes just one piece of the jigsaw puzzle. The whole of life - every aspect of our experiences - has to be drawn into what I call: the God experiment.

From the book: