A Critique of the Science in Russell Stannard’s “The God Experiment”
Review of Russell Stannard’s The God Experiment: Can Science Prove the Existence of God? Mawhah, NJ: Mindspring. 1999.
The Prayer Experiment
Stannard’s overarching view is that science and religion have very much in common. Science is a little more experimentally ambiguous, tentative and broadly inferential than we commonly think and theology is more empirically approachable. They have quite similar methodologies; they make continual progress over time; they share insights which are mutually beneficial, they even cross fertilize each other’s ideas.
He devotes his opening chapter to a major experimental realization of this view. It illustrates the deep dilemmas just beneath the surface of this kind of melding of science and religion that are easier to gloss over in the more abstract discussions.
The chapter describes a current multi-center trial of the efficacy of prayer in enhancing the surgical outcome for 600 coronary artery surgery patients, results due in 2002. There is a body of report now in the literature of studies like this though not of this scale. The implicit message of this and related studies is, ‘We are just out there doing the science like any other regular scientists and letting the die fall where they may.’
This is a large and expensive study and obviously of serious intent at least in terms of the effort and monies invested. It is centered at the New England Deaconess Hospital at Boston. Stannard is a trustee of the foundation funding the study and it is illustrative of his central thesis – the exploration of the extent to which theology can be regarded as a science.
Something very different to normal science is going on here and I do not just mean the investigation of a supernatural effect. At first glance it looks like a large scale experiment in an area where the preliminary work suggests something interesting might be happening. Something interesting of course would be a gross understatement as what is being suggested is the reproducible measurement of supernatural phenomena. It is not just a routine study like many thousands of others which map psychosomatic functions such as, ‘Having a positive outlook, a strong set of personal beliefs or going to church etc can improve medical outcomes.â€™ He is explicit – the study aims to specify a supernatural outcome. Fundamental issues underlie this experiment and reflect on the whole field of the current science/religion debate.
The Design of the Experiment.
We can get a good handle on what is really going on in the prayer experiment if we simply take it at face value – consider it as one would a regular scientific experiment and see how it looks.
The first oddity is the experimental design. The main effect chosen is very complex to measure – the long term outcomes post coronary artery surgery. It is also a hugely expensive trial.
For revolutionary science dramatic evidence is required. If it is intended to scientifically demonstrate miraculous healing there are far less ambiguous ways to do so and there is no need to go to anything like this expense. One could have people pray for the miraculous reattachment of cardiac mitral valve leaflets torn off their attachments to the heart muscle after a heart attack. It is easy to follow what is happening sonographically, and there is an unequivocal outcome. The leaflets don’t reattach themselves in any natural way. It is either surgery or a miracle.
Alternatively one could pray for a vital organ like a kidney to be provided to neonates born without them – very simple to specify and measure, life threatening in the extreme, and an unambiguous outcome. You wouldn’t even need a control group to impress everyone and very few patients would be needed.
The biggest post coronary surgery risks are bleeding and stroke. Moderately dangerous and sometimes catastrophic post operative bleeding are the top serious risks. Prayer for the miraculous alteration of this course of events is a request for a very substantial intervention. Replacing the damaged platelets in the blood stream is the most frequent thing that surgeons do to treat post operative bleeding. Supernatural intervention at this level is a major biological event no less impressive than the more readily measurable events suggested above.
The big difference between the coronary trial and the very simple low cost experiments above is the ease with which one can monitor what is going on. A positive outcome might be interpreted as divine miraculous intervention, though an alternative explicit hypothesis might be that there is some parapsychological phenomenon like telepathic healing going on. The next oddity is that there is no control for this. It would have been very easy to include such in the design and rule it in or out as a competing explanation of a positive effect.
It is difficult to understand why this would not be done when claims of such magnitude are under consideration. A claim that merely the power of prayerful activity is being tested without interest in discriminating the mechanism when such hugely different contentious mechanisms are competing explanations would seem very strange and would normally be thought a major error of experimental design in a scientific study like this. This is particularly so since the experimenters know that a positive outcome would be widely used as a support for the reality of miraculous healing.
Science is about putting one’s hypothesis on the line, designing the simplest most powerful experimental test and living with the consequences. Whilst it is true that scientists will sometimes hang on to their favorite hypotheses in the face of initial negative experimental outcome the purpose of the experimental program is to be the theory deal breaker. This is an experiment with no possible negative outcome. A negative outcome will not be taken as evidence against miraculous healing even by the experimenters. God may fail to cooperate etc. The claim is that a positive outcome will not be an unequivocal demonstration of a miraculous process as there is an uncontrolled alternative hypothesis. Everyone knows in advance of course how a positive outcome will be received. Stannard says he is a little bemused at all the fuss and attention the study is getting. Clearly the journalists calling him are aware of the implications of what is going on and the way a positive outcome will be universally interpreted.
The next question that might be asked is how the main effect is produced and specified.
In the coronary study the people praying for the patients are given only the first names of the patients they are praying for. With six hundred patients and (?) six hundred controls presumably there are many patients named John. It would be very important in this confusing circumstance to be very specific in the instructions to intercessors. How would the intercessors be instructed to specify their prayers?
‘Please God intervene miraculously in the healing process of the John that I have been allocated (and no others) in the experimental group not any of the Johns in the control group.’
The John who is to be saved, by the way is in the experimental group by chance and not because of some virtue, unless God is rigging the allocation to experimental and control groups. Presumably God then performs a miracle on the experimental John altering the outcome of the surgery.
The above account is not meant to offend in any way. But inevitably it starts to sound offensive because the proposition becomes manifestly absurd when examined closely. Any deep or detailed description of what one thinks is going on sounds like the above. Do any groups no matter how anthropomorphic in their god beliefs really think that this is how the world is? Do gods who make universes sit still for this kind of micromanipulation?
Consider further the way a presumed experimental effect like this would be routinely defined in further trials.
Suppose the experiment is successful, it is repeated and the effect is reliable and a second experiment rules out telepathic healing as an explanation. We might next choose representatives from various religious traditions (and also atheists perhaps as controls) to see who are the most effective at intercession. We should vary the number of people praying to see how many patients the intercessors can effectively pray for. We should see what kind of prayer work best – prayers of praise – beseeching prayers – prayers of penitence and so on. This is the road of science.
We could establish whether the intercessors even need to know the patient’s name. We should investigate whether there any very gifted folk who can bring about the miraculous cures of thousands perhaps through their intercessory prayers? They would be very important to identify.
Somewhere down the track we would arrive at an outcome just like any other equivalent experimental program.
The outcome could look like this. Lutherans and Mormons, are demonstrated to be the only effective Christian groups in praying for miraculous healing. One Islamic fundamentalist intercessor is also successful and both Hindu and Buddhist representatives have failed. On average 5.6 patients can be effectively prayed for by one experimenter before the effect falls off. Prayers for cardiac healing are twice as effective as prayers for breast cancer healing. It takes 2.3 days on average to get a mitral valve spontaneously repaired.
What is wrong with the above, apart from the fact that it sounds blasphemous? Again it is not intended to offend. This is exactly how a real experimental program would be pursued. What is wrong is that no one, not the experimenters, not the most loyal of religious constituents believe that the world is like this.
The reason no one is concerned that we would end up in the above absurd scenario is that this project is not about the start of a detailed, on going scientific investigation which could in fact produce such an outcome. This is not about science; it is about using science as a prop for a belief system. We rightly protest when science is used to prop up ideological/political views such as the race/IQ pseudo research. This experiment may seem innocuous but in the end it is bad for religion and bad for science.