Why Historians Cannot Confirm Miracles: The Limits of Historical Methodology

Religious apologists often maintain that miracles not only occurred in the past but that historical methods can support this claim. This article argues that they cannot.

The Role of History

We have no direct access to the past. Events that once unfolded in real time are gone forever, surviving only through scattered documents, artefacts, and,  in rare cases, the memories of living witnesses. The historian’s task is to reconstruct, as accurately as possible, what most probably happened, using the available evidence together with our broader knowledge of the period. Although certainty is unattainable in history (as in any academic discipline), historians can sometimes establish certain events or the existence of individuals as historical facts. In other cases they express conclusions in terms of probability, usually through qualitative language rather than numerical values, with the degree of confidence reflecting the strength of the evidence. When evidence is absent or insufficient, the honest response is simply that we cannot know what happened. Importantly, historians do not claim the event did not occur; such a negative assertion would itself require evidential support. Like scholars in any field, historians strive to draw conclusions as objectively as the evidence allows.

Criteria for the Best Explanation

Historians employ various methods to determine what most likely happened. One widely used approach is inference to the best explanation. When the evidence strongly satisfies the relevant criteria, historians may justifiably accept the resulting conclusion as reliable. C. Behan McCullagh, in Justifying Historical Descriptions [1], summarises the main criteria as follows:

1. Explanatory scope — the hypothesis must account for a greater variety of data.

2. Explanatory power — the hypothesis must render the data more probable than rival hypotheses.

3. Plausibility — the hypothesis must be implied (to some degree) by a wider range of accepted beliefs than any rival, and implied more strongly; its negation must be implied by fewer beliefs and less strongly.

4, Ad hoc — it must introduce fewer new suppositions about the past that are not already implied by existing beliefs.

5. Disconfirmed by fewer beliefs — it must conflict with fewer accepted statements.

6. Falsifiability — it must be testable to a greater degree than rival hypotheses.

Criteria 1 and 2 depend heavily on the specific evidence and can be set aside for present purposes. Although many critical scholars already question the reliability of the gospel accounts, that issue is not central to the present argument. Several of the criteria bear on miracle claims,  especially falsifiability (6), but here I focus on plausibility (3). Assessing plausibility can be difficult. For example, if a newly discovered speech is attributed to a well-documented historical figure but contradicts everything reliably known about their character, values, and habits, the historian will be sceptical. When such incongruity is supported by independent evidence, the attribution is usually rejected. Of course, people sometimes act ‘out of characte’”. so such judgements must be made with care.

Miracle Claims

Gospel miracle accounts, including the resurrection, typically follow this pattern:

1. An event is said to have been witnessed by one or more reliable observers.

2. Later writers or apologists argue that the event cannot be explained by natural causes.

3. The conclusion is drawn that a miracle must have occurred.

In antiquity, events we now classify as miraculous were often regarded as part of the normal order of things; the ancient reports themselves rarely label them “miracles.” The claimed testimony can be substantial — hundreds of witnesses in the case of the early resurrection creed [2], thousands for the feeding of the multitudes [3], or large crowds in modern examples such as the Miracle of the Sun at Fátima or healings at Lourdes [4]. Written accounts appear later, and their authors frequently assert a miraculous cause, often without rigorous critical scrutiny, even when natural explanations are available. In extreme cases, such claims have led to violence or exile, as in the 2005 Mumbai crucifix incident [5].Apologists then argue that the volume of testimony, combined with the supposed impossibility of natural explanations, makes the miracle the best (or only) explanation. In the case of the resurrection, many Christians go further, asserting that the evidence is so strong that the event qualifies as a historical fact.

From the perspective of historical methodology, this reasoning fails on two grounds. First, most critical scholars do not regard the gospel narratives as historically reliable. Second, and more fundamentally,miracles cannot be incorporated into historical explanation. This second point is the focus of the remainder of the article.

Miracles and Historical Plausibility

To claim that a miracle occurred is to make an assertion about the way the world operates. While ‘miracle’ can mean merely an extremely improbable event, I use it here in the stronger sense: an event that violates or suspends the laws of nature (as in walking on water, turning water into wine, or rising from the dead after three days).Returning to the criteria for the best explanation, the decisive issue for miracle claims is plausibility (3). Are such events believable in light of everything we know about how the world functions, especially the world of the first century? The answer is clearly no. We have no confirmed experience of such events, nor any verified mechanism by which they could occur. To assert that they did happen is to introduce new, unverifiable knowledge about the operation of nature.

Some apologists object that this stance merely reflects a naturalistic worldview or methodological naturalism that begs the question against the supernatural. Philosophers like William Lane Craig argue that Hume's classic argument (uniform experience rules out miracles) is circular: it assumes miracles have never occurred to conclude they cannot occur. Others, including Tim and Lydia McGrew, propose Bayesian approaches where strong testimony could outweigh low prior probabilities if a cumulative case (e.g., multiple independent facts) is built, and if theism is sufficiently probable on other grounds. They contend historians can assess miracle claims without a priori ruling them out, as long as evidence is weighed neutrally.

This criticism has force if one already accepts a high prior probability for divine intervention (e.g., via independent arguments for God's existence). However, history, like the natural sciences, seeks conclusions that do not depend on any particular metaphysical commitments. To say a worldview ‘allows’ miracles is not a statement about current human knowledge; it is an assertion that events can occur outside the natural order and interact with it. Appeals to the existence of God (“if God exists, miracles are plausible”) simply beg the question: they assume the very entity whose action is being proposed as the explanation. Moreover, historians of other faiths or of no faith could never accept a conclusion that presupposes a specific deity. Historical claims based on observable, verifiable evidence must, in principle, be acceptable to any competent historian regardless of personal beliefs.

Example

A hypothetical case illustrates the principle. Suppose we have excellent, independent evidence that Julius Caesar was in Gaul on 15 November 52 BCE and equally strong evidence that he was in Rome on the same day. Numerous reliable witnesses attest to seeing him in each location, despite the roughly 1,500 km distance. What would a historian conclude? Not that a miracle occurred. Even with impeccable testimony, the historian would note the anomaly and conclude, “We do not know what happened.” No known transport system of the time could have covered such a distance in a day, so the idea that Caesar did so is simply not plausible on our knowledge of the period.

Now consider a modern parallel: Phil Collins is reliably reported to have performed at Wembley Stadium in London and, on the same calendar day, at John F. Kennedy Stadium in Philadelphia. The testimony is excellent and the event is accepted as historical fact. Why the difference? Because Concorde existed, supersonic flight was possible, time zones allowed it, and the logistics are entirely feasible given the technology of the 1980s. The explanation is therefore plausible and grounded in the known realities of the time.

Conclusion

In the world we inhabit, past or present,we have no experience of dead people returning to life or of any other events that suspend natural laws. Until such evidence appears, historians cannot responsibly propose miracles as historical explanations; they lack both the tools and the methodology to do so. It would be comparable to asking a geologist to determine the charge of an electron: the discipline can use relevant findings from other fields (e.g., radiometric dating), but it cannot itself discover new principles of physics.

Should compelling evidence eventually emerge that certain types of “miraculous” events are possible, that evidence will come from medicine, physics, or other scientific disciplines, not from historical methodology. Historians would then reassess relevant texts in light of the new knowledge. That is simply how historical inquiry proceeds.

Anyone is free to believe in religious miracles on theological, philosophical, or personal grounds; no one can prevent that. Such belief, however, cannot be supported by the historical method.

Notes

1. McCullagh, Justifying Historical Descriptions 1984

2. I Corinthians 15:5-8

3. Luke 9:12-17, Matthew 14:13-21, John 6:5-15, Mark 6:30-44, Matthew 15:32-39, and Mark 8:1-9

4. José Barbosa Machado, The Miracle of the Sun 2012, John Lochran The Miracle of Lourdes 2008

5. Christian Today 14 November 2016

August 2021