This past week, in the course of a discussion concerning the resurrection of Christ, Nicholas Covington borrowed an argument from philosopher Matt McCormick in an attempt to show that eyewitnesses to alleged miracles are very, very unreliable. The focus of this post is on slide 2 from McCormick’s presentation, where he makes the following assertions:
- Assertion 2.1: 8,000,000+ pilgrims have visited Lourdes in over a century.
- Assertion 2.2: “Suppose, charitably, that half experienced something they took to be supernatural.”
- Assertion 2.3: “66 miracles have been declared to be real by the official investigating body of the Catholic Church.” Note: as of this writing the number of officially recognized cures at Lourdes is 67 (which McCormick gets right on slide 8).
- Assertion 2.4: “Provisional miracle testimony reliability = 0.0000165”
Rebuttal 2.1: In Assertion 2.2 McCormick appears to be guessing at how many people have claimed to experience miraculous cures at Lourdes. He provides no reason to believe his guess is at all accurate. Note how his guess is wildly different than the estimates cataloged by Craig Keener:
By some estimates, most groups of organized pilgrims in the first half century reported between 0.001 and 0.02 percent cures . . . . (678)
Of millions of pilgrims over the years, with probably more than half a million sick, about five thousand miracle claims have been recorded (less than 1 percent). Of these, the medical committees evaluating cures there have judged a smaller number to be “medically inexplicable.” Only these screened cures have been forwarded on to bishops for further discussion, and of these only sixty-five have survived the evaluation of all examiners and been pronounced certain “miracles.” If the only genuine cures are those that have been deemed both medically inexplicable and religiously approved, the cure rate is abysmally low — perhaps no better than one would find in hospitals dealing with the same number of “incurable” patients. Even many who affirm that supernatural healings occur there do not contend that they should be expected very often. (679)
McCormick’s over-estimation is particularly ironic considering that on slide 4 he writes, “their [the disciples’] error rate with regard to supernatural claims would be high: they would conclude that miracles are more common than they really are, for example.”
Rebuttal 2.2: McCormick’s guesswork in Assertion 2.2 is charitable towards the skeptic not the believer. The value in Assertion 2.4 will approach zero as the number of miracle claims goes up.
Rebuttal 2.3: The mere fact that the Medical Bureau does not officially declare a miracle claim to be a true miracle does not mean that the witnesses were incorrect or lying about what they saw. On the one hand, they may be accurately describing a cure but the cure was natural instead of supernatural. In this case they were incorrect in interpreting the event but accurate in describing what they saw. On the other hand, a supernatural cure could occur but the claim does not pass certain standards put in place by the Medical Bureau. The number of verified miracles merely provides us with a lower bound in calculating the provisional miracle testimony reliability (at least if you are going to accept the results of the Medical Bureau). Craig Keener explains the rigorous standards at Lourdes:
When one examines the statistics, though, one should also examine the stringent criteria employed there — cures are more frequent than the critical minimum finally counted as miracles. The concern of many pilgrims is their travel and health, not obtaining documentation; to satisfy the Medical Bureau’s investigation, however, they must have full medical documentation of their prior condition, including any available hospital records. Investigators cannot always obtain sufficient documentation to proceed, since many doctors do not respond to requests for it, lacking interest and/or time. Tracking down the witnesses can be difficult if they have moved. If patients have taken medicine that could possibly have produced the cure, they are excluded from consideration. Most patients today, however, will seek the best of medicine as well as prayer. If one’s ailment is not demonstrably organic, one is excluded, even though this excludes from consideration many genuine afflictions.
Many people find themselves cured, yet if there cures cannot pass all the stringent criteria to achieve certainty, they are not pronounced miraculous by the church. For example, a Protestant doctor notes that he “saw with my own eyes” that a woman who had been suffering from “advanced ankylosing spondylitis of the spine” was now able to pick up objects from the floor “without the slightest pain or difficulty.” Lourdes rejected this cure because X-rays showed that her spine remained diseased, but the doctor notes that her new ability defied scientific explanation. More than twelve hundred cures accepted by the Medical Bureau after multiple examinations lacked some (though usually not all) of the requisite documentation and thus were never processed. Anyone who has tried to collect medical documentation recognizes that factors beyond the researcher’s control sometimes make this process difficult. Besides the fairly secure cures, the bureau knows of four thousand other probable cures. Many others were cured and never submitted any information to the bureau. Forty doctors confirmed the cure of a medically incurable, quadriplegic postencephalitic idiot — a child who went from complete insensitivity and lack of control to intelligent normalcy — but because some documents were missing, the cure was never officially proclaimed. By the minimalist standards employed at Lourdes, it is likely that many of Jesus’s miracles reported in the Gospels would have been screened out had they occurred instead at Lourdes.
Unfair skepticism at any point in the process, whether from doctors or church leaders, can skew the results just as credulity throughout the process would. The makeup of the committee and their philosophic predispositions, as well as those of bishops and diocesan councils, apparently plays a noteworthy role; some eras and regions include a number of reported healings, whereas others report barely any. For example, no healings were admitted “between 1913 and 1946.” Yet between 1925 and 1946, thousands of reported cures appear in the Medical Bureau’s records, of which eighty-nine were viewed as inexplicable; those responsible for evaluating them simply did not proclaim any of them as miracles during this period. After the long hiatus, twenty-four miracle proclamations cluster in thirty-four years, starting, as one writer observes, “at about the same time as the reorganization of the Medical Bureau and the decisions to encourage the simultaneous development of the Marian shrine of Fatima. The French church decided to acknowledge healings at Lourdes near the fiftieth anniversary; they proclaimed twenty-one cures, including some that had taken place earlier, over the span of a few months, a collection that even today remains roughly a third of those proclaimed there. That is to say, discerning an individual miracle can involve subjectivity, whether in affirming or denying it. (679-681)
It is intriguing to see how one could play with the numbers: ((67 certain cures) + (4000 probable cures)) / (5000 cure claims) = 0.8134. I’m not claiming that these numbers are exact (e.g., in another place Keener  says there are roughly 6000 claims). My point is that more careful estimates of the numbers from Lourdes indicate that eyewitnesses are not even close to being as unreliable as McCormick believes.
Keener, Craig S. Miracles: The Credibility of the New Testament Accounts. Baker Academic, 2011.