Public theology and prophecy data: Factual evidence that counts for the biblical world view

Journal of the Evangelical Theological Society, Mar 2003 by Newman, Robert C, Bloom, John A, Gauch, Hugh G Jr

[It] is worth noting that successful prophecy could be regarded as a form of miracle for which there could in principle be good evidence. If someone is reliably recorded as having prophesied at t^sub 1^ an event at t^sub 2^ which could not be predicted at t^sub 1^ on any natural grounds, and the event occurs at t^sub 2^, then at any later time t^sub 3^ we can assess the evidence for the claims both that the prophecy was made at t^sub 1^ and that its accuracy cannot be explained either causally (for example, on the ground that it brought about its own fulfilment) or as accidental, and hence that it was probably miraculous.

J. L. Mackie1

I. INTRODUCTION

The key feature of prophecy which allows it to provide humans with evidence that God exists is that successful predictions, though hard to make, can be easy to check. The "hard to make" part reveals God's involvement, while the "easy to check" part enables human discovery. Prophecy has content "which God alone can know at the time of the revelation," since it is "not predictable from natural laws" accessible to human understanding, yet this content is the sort "which humans can certainly discover afterward."2

As noted in our previous paper, for prophetic data to be admissible as evidence that can count across world views, each prophecy must satisfy four criteria: (1) clear prediction; (2) documented outcome; (3) proper chronology; and (4) evidential weight. In addition, the dataset as a whole should satisfy another four criteria for the data to be relevant: (5) testable hypotheses; (6) world view import; (7) robust conclusion; and (8) manageable effort. These criteria unpack as follows.

(1) Clear prediction. The prophecy must be publicly available with a reliable text and evident interpretation. Its predictions must be sufficiently specific and detailed that a fulfillment, and also a failure, would be recognizable without any ambiguity.

(2) Documented outcome. The outcome of the prophecy must be evident by the present time, with that outcome well documented by publicly available facts. For instance, reliable and independent historical records count, as do the stones and relics found at archaeological sites and museums. Evident facts of world history also count. But unverifiable reports do not count, especially reports of miraculous events that are exceedingly improbable from atheistic or other perspectives.

(3) Proper chronology. Definite empirical evidence must be publicly available to document that the prophecy predates its fulfillment. For the OT, this criterion is satisfied by all outcomes dated after 150 BC, the average date of copies of Bible books among the Dead Sea Scrolls. This is also about the time the independently-circulated Greek translation, the Septuagint, was completed in Alexandria, Egypt.

Likewise, for a collection of books such as is found in the Bible, the corpus or canon must have been settled before the considered outcomes began. Otherwise, knowledge of the outcomes could have influenced the selection process, canonizing those books with fulfilled prophecies while discarding other books with embarrassing ones, thereby producing a spurious prophetic accuracy using the unfair advantage of hindsight.

(4) Evidential weight. Predictions must be sufficiently specific and unusual to make their fulfillments unlikely merely by chance. For instance, a generic curse that a city will be destroyed has little evidential weight because most ancient Near-Eastern cities have been destroyed many times. Furthermore, there must be factual reasons for assigning particular odds of fulfillment by chance, such as 1:5 in one case or 1:100 in another. For instance, the antecedent odds for a city encountering some particular outcome can be assigned by determining the proportions of the various possible outcomes for a sizable and representative reference class of comparable cities. Sometimes simply counting the number of antecedently equally probable outcomes can provide a satisfactory assignment.

(5) Testable hypotheses. Hypotheses are testable when they make different predictions about some observable outcome. The expectation of the Christian world view for the Bible prophets is high accuracy. Consequently, any other world view that expects a markedly lower accuracy has thereby rendered the Christian and that other world view testable.

(6) World view import. These different predictions, such as high or else low prophetic accuracy, must originate from causal explanations with significant world view import. For instance, in a competition between Christianity and naturalism, the causal explanation for the prediction of high accuracy is that God alone knows the end from the beginning and has revealed the future to prophets, whereas the causal explanation for the low accuracy is that humans (and more generally any physical systems) have severely limited predictive powers, although occasional lucky guesses are expected.

(7) Robust conclusion. The verdict on the Bible prophets' claims of predictive accuracy must emerge from major and settled features of the data, not from picky and disputable details. Different persons with different data subsets, different analyses and interpretations, and even vastly different expectations originating from diverse world views, should all reach virtually the same conclusion. Two properties that greatly favor robust investigations are that the inquiry's data produce an evidential weight rising exponentially with the amount of data and that the inquiry's analysis is disentangled from other information and world view beliefs.