Arguing the Affirmative: RANDAL THE CHRISTIAN
Arguing the Negative: JOHN THE ATHEIST
Randal’s Opening Statement
Archbishop of Canterbury William Temple once observed, “When I pray coincidences happen, and when I do not pray coincidences do not happen.”[70] Many Christians can resonate with Temple’s wry reference to God’s providence. But atheists demur, charging that such experiences only evince a selection bias that counts the hits and ignores the misses. So who is right? To answer that question we’ll consider a specific case from a personal friend of mine, Kent Sparks, Professor of Biblical Studies at Eastern University.
Kent was a pastor living with his wife, Cheryl (a physical therapist), in North Carolina when they were pursuing an adoption through a crisis pregnancy ministry named House of Ruth in Downey, California. But since they had not had success with House of Ruth over the previous year and a half, they proceeded to close on a private adoption in Georgia. After they closed on the adoption of their daughter Emily and returned to North Carolina, Kent called House of Ruth and left a message requesting the agency to suspend their file. Little did Kent know that at that very moment the House of Ruth staff was in a meeting with a young woman who chose their file for her child. As soon as the meeting ended, a staff member called the Sparks to inform them of the good news. Cheryl answered the phone, assuming they were returning Kent’s call. Needless to say, she was shocked to learn instead that they were calling to offer a second child for adoption! Overwhelmed by the prospect of accepting a second infant, Cheryl called a friend to ask for prayer. Later when Kent arrived home from work, Cheryl asked him to conduct a family devotion without informing him of the situation. Perplexed, Kent opened his Bible and read from Proverbs 3:27: “Do not withhold good from those to whom it is due, when it is in your power to act.” Shortly thereafter Cheryl’s friend called her and said, “I have a verse for you.” She then quoted Proverbs 3:27. Based on that collocation of events, Kent and Cheryl accepted the adoption and welcomed their second daughter, Cara, into the family.
Is it reasonable for Kent and Cheryl (and us) to believe that this adoption was divinely endorsed? In order to answer this question we should consider the concept of a design filter.[71] When we seek to identify agency as an explanation for an event, we first seek to establish that the event was contingent. Thus, an event that is a known result of natural law is not explained via design. (That’s why we don’t invoke Jack Frost to explain the frost on our window panes.) Next, we need to eliminate the possibility of chance. We do that by looking for events that are sufficiently complex and specified to a situation. If the event is contingent, complex, and specified, then a design explanation for the event is warranted.
To illustrate, imagine that Fred Smith passes a movable type sign every day on his drive to work. On the day of his fortieth birthday the sign says “Happy Birthday!” Could Fred reasonably believe that the message was put there to wish him a happy birthday? That depends. The first hurdle to clear is an analogue of contingency. For example, if the sign always said “Happy Birthday!” then no design inference would be warranted. But if this was the first day that the message appeared on the sign, then the event could be considered contingent. From here the more complex and specified the message, the stronger the evidence is that it’s for Fred. Thus a sign that reads “Happy Birthday Fred!” offers better evidence, while one that reads “Happy Fortieth Birthday Fred Smith!” offers excellent evidence.
Kent and Cheryl’s case evinces these same hallmarks of contingency, complexity, and specification. While these events are obviously contingent, they are also complex since they involved multiple factors timed together (e.g., Kent’s call concurrent with the adoption meeting), and they included specified information (e.g., two independently confirmed references to Proverbs 3:27). Consequently, Kent and Cheryl (and we) are fully justified in drawing the conclusion of divine action in confirmation of the adoption.
Do things change if we also count the misses? No. Fred may have never before seen a message on the sign relevant to him, but on the one day that it says “Happy Fortieth Birthday Fred Smith!” he justifiably believes the message is for him. Likewise, even if Kent and Cheryl have never had another experience like this, the contingency, complexity, and specification of these events are sufficient for the Sparks to believe God affirmed Cara’s adoption.
John’s Opening Statement
Almost every scientific study done on prayer has shown that prayers are not statistically answered any better than luck.[72] Research has shown us that people are prone to misjudge the true probabilities for any given event—we’re often wrong. Take a deep breath right now. Mathematician John Allen Paulos asks us to consider the odds of whether we have just inhaled a molecule that came out of Caesar’s mouth when he said, “You too Brutus?” You’ll probably doubt this, but he shows that the odds of this happening stand at about 99 percent.[73]
The fact is that incredible coincidences are common, even virtually certain, given enough opportunities for them to occur in the lives of millions of believers. The most we can say about them is that their causes are unknown. So once again Randal is arguing from the gaps—a known, informal fallacy. We should not trust personal anecdotal evidence when it comes to answered prayers, especially since believers count the hits and discount the misses due to a thousand qualifications. Besides, believers in each era will only pray for things they expect can happen, and what they expect depends on the state of contemporary science. Still, the question remains why they have to ask their God for anything at all if he really cares for them as a mother cares for her children. I see no reason why they should.
Even if Christians receive an answer to a prayer, this doesn’t show their particular God exists, since a different god may have answered it out of compassion. And if in reverse, the Christian God is answering the prayers of other believers in hell-bound false religions out of compassion, then God is helping to send them to hell by providing confirming evidence for their faith against Christianity.
When it comes to miracles due to prayer, it gets worse. Believing in them demands a nearly impossible double burden of proof. What believers must show is that a miracle could not have happened within the natural world because it was nearly impossible (or else it’s not considered miraculous). Then they must turn right around and claim such an impossible event probably took place anyway. The probability that a miracle actually took place is inversely proportional to the probability that such an event could take place (i.e., the less probable it is that a miracle could take place, then the more probable it is that it didn’t take place), so the improbability of a miracle claim defeats any attempt to show it probably happened. That’s why extraordinary claims of miracles demand a greater amount of solid evidence for them (e.g., if a person tells us he levitated, we need more than just his testimony to believe him).
There is one way to evaluate the reliability of human testimony to miracles, and it comes from the findings of the Catholic Church. Dr. Matt McCormick asks us to consider the millions of miracle claims alleged by believers in Lourdes, France:
The Catholic Church has officially recognized sixty-seven of them. A rough estimation of the general reliability of human miracle testimony from Lourdes comes out to be a mere .0000167. That is, in general, when humans give miracle testimony, their reliability is orders of magnitude worse than it needs to be for us to even provisionally accept it.[74]
While I would want to investigate personally the sixty-seven cases this believing church concluded were miracles, it still shows that we have every reason to be skeptical when believers claim a miracle took place. Moreover, since I have never seen one, I am within my rights to doubt them all. Believers have built in biases to accept miracles because they have a need to believe them. It basically shows us that people are still agency detectors—something we inherited from our animal ancestors. Since we know this about ourselves, it should cause us to be skeptical that there are agents behind unexplained events.
That there are some unexplained events I’ll admit, just as there are some good things said in the Bible concerning women and animals. But the bad things must be explained and not just explained away.
This raises the problem of suffering (or evil). There is no excuse for a good God not to do miracles for the millions of believers who suffer intensely around the world. For every personal claim of a miracle healing, there are perhaps hundreds of thousands of people who receive no such thing.
Randal’s Rebuttal
John avers that “incredible coincidences” occur on a regular basis. True, and the design filter screens out events if they are merely improbable. It only triggers when an event is highly improbable and specified to the situation, like the sign wishing Fred a happy fortieth birthday or the collocation of events confirming that the Sparks should adopt Cara.
John protests that events like this don’t establish that the deity of one’s personal faith caused them. This is surely a desperate point. If I pray to the Christian God and the prayer is answered, why should I think any other being answered it? Finally, John claims that a Christian can only infer divine action if no possible natural cause could have produced the effect. This is false. A chain of chance natural causes could have spelled out “Happy Fortieth Birthday Fred Smith!” on the sign, but the most plausible explanation is still an intelligent cause. John has utterly failed to establish that one cannot reason similarly in the cases of answered prayer.
John’s Rebuttal
What Randal describes is personal anecdotal evidence that no scientist would ever accept as evidence, since incredibly improbable events happen to people all of the time. They are even virtually certain to happen to someone sometime given enough opportunities for them to occur in the lives of millions of believers. The really surprising thing isn’t that these events happen but that they don’t happen more often.
How we test prayer scientifically is what the American Heart Journal did when testing for answered prayer on patients who had heart bypass surgery.[75] The patients were separated into three groups. Group 1 received prayers and didn’t know it. Group 3 received prayers and did know it. Group 2 received no prayers and didn’t know it (science must have a control group like this representing the null hypothesis). Groups 1 and 3 were prayed for by various congregations throughout America. The results were very clear. There was no difference between the patients who were prayed for (groups 1 and 3) and those who were not prayed for (group 2). Moreover, the patients who knew they were being prayed for suffered significantly more complications than those who did not know they were being prayed for.
Randal’s Closing Statement
What John dismissively refers to as “personal anecdotal evidence” is vetted testimony—a type of evidence that is treated as of great value in a court of law, so why not here? Instead of addressing the logic of the design filter, John bewails my alleged failure to count misses. That’s an ironic complaint since his arguments prevent him from ever counting hits.
John’s Closing Statement
We do indeed have a design filter preprogrammed into us from our animal ancestors to see agents behind improbable events given the proper circumstances. We are also not good at predicting the actual probabilities of an improbable event. So we should not trust personal anecdotal evidence but instead trust the scientific studies on prayer.