All historical argumentation is probabilistic. This is also to say that any and all historical hypotheses are subject to revision or dispute. Hypotheses subject to revision are hypotheses whose probability sufficiently approaches 1.0 that we can treat them as virtually certain. Such hypotheses include the hypothesis that Germany invaded Poland in September of 1939, or that Jesus of Nazareth existed. Such hypotheses are virtually certain not necessarily because there are no conceivable alternatives, but in many (perhaps most) cases because all conceivable alternatives are sufficiently improbable that they can be excluded. Can I conceive of a world in which all the documentary and eyewitness evidence for Germany’s invasion of Poland in 1939 is falsified and it never took place? Perhaps. Is that alternative probable? Hardly. Nonetheless, in principle, even the most probable statement is subject to revision upon the emergence either of new evidence or new insights into old evidence. The recent resurgence in arguments for Jesus’ historical non-existence rested entirely upon the argument that there had emerged new insights into old evidence. The reason that these arguments fail is because those competent in the matter and fully familiar with the evidence recognized immediately that these were not new insights at all but almost without exception insights that had been advanced and rejected the better part of a century ago. For instance, much of the argumentation rested upon a literally Victorian-era understanding of the nature of ancient myth-making, which has long since and properly been abandoned as empirically unsustainable. There is a reason that one can count on two fingers the number of credentialed New Testament scholars who subscribe to the hypothesis that Jesus never existed: quite simply, competent familiarity with the data precludes affirmation of the hypothesis.
Hypotheses subject to dispute are different. These are hypotheses with which a competent person fully familiar with the evidence can reasonably disagree. For instance, it is a virtual certainty that Jesus died on a cross sometime around Passover, between the years 29 and 35. I would argue that among those years, 30 (the long-time majority opinion) remains the most likely. I can present a number of arguments in favour of that date. However, I recognize that a competent person fully familiar with the evidence could reasonably argue for any other date within that range. As with the case of virtually certain hypotheses subject to revision, new evidence or fresh insights into old evidence can alter the probabilities. For instance, the discovery of the Delphi inscription a century ago has significantly narrowed down the date for Paul’s meeting with Gallio as attested in Acts 18, such that what was once thought to have occurred sometime between the late-40s through the early-50s can now more precisely be said to have occurred sometime between July 1st of 51 and June 31st of 52. This range can now in fact be treated as a virtual certainty. What remains subject to dispute is when within this range the meeting took place. For instance, it can be and has been argued that the meeting is more likely to have taken place very early in that range, on the grounds that Paul’s opponents very conceivably seized upon the change in governor in order to gain a hearing for their charges; this is quite plausible, and yet can hardly be treated as a virtual certainty and is such that a competent person fully familiar with the evidence could reasonably disagree.
Both the year of the crucifixion and the timing of Paul’s appearance between Gallio demonstrate that the heuristic distinction between hypotheses subject to revision and hypotheses subject to dispute allows for a dynamic understanding of probability. In my own current work—i.e. my long-percolating study of and obsessive concern with the dates at which the New Testament texts—hypotheses subject to revision tend to yield relatively large ranges. It is beyond reasonable dispute that the Gospel of Mark was written sometime after Jesus’ death and before the first attestation to the gospel’s existence. This allows for a date anytime from c. 30 to c. 120, the latest likely time at which Papias writes about Mark’s compositional work. Dates outside this range are almost certainly non-starters, although the higher end is more likely subject to revision than the lower as the probability that Mark’s Gospel was written before Jesus’ death approaches zero (although in principle one could not rule out the possibility that parts of the text were written during his lifetime). Early 21st-century scholarship has tended to favour a date about midway through this range, with c. 65 to 75 probably representing something like a majority opinion, but the salient point here is that all debates regarding where in the range c. 30 to 120 to situate Mark’s Gospel have moved from matters of revision to matters of dispute, because we are now arguing about matters upon which competent persons fully familiar with the evidence can reasonably disagree. Awareness of this heuristic distinction between hypotheses subject to revision and hypotheses subject to dispute allow us to avoid wasting time and energy either disputing that which is beyond reasonable dispute or seeking virtual certainty when the data does not allow us to do so.