Why Richard Carrier probably won’t win over the guild, part 2

Chaplin in Modern Times

Bayes isn’t as mechanical as some seem to think.

How confident should a reasonable person be that Jesus was a real man who actually lived? Academic biblical scholars, “the guild,” display little doubt. Dr. Richard Carrier urges a skeptical view in his book, On the Historicity of Jesus.

The first post in this series (link) specified two alternative hypotheses about Christian origins taken from Carrier’s work. The next step in a Bayesian-style analysis is to compare the plausibility of the contending hypotheses based on general background information, before examining how more specific evidence about Jesus’ historical or mythical status might change the relative plausibilities.

Step 2: assessing an initial plausibility from background information

Bayes theorem governs how new evidence changes one state of belief into another. Bayes says nothing about what the very first state of belief in an investigation should be, only how evidence shapes later states from whatever that first one was.

One popular approach to formulating an initial belief state is to compare how relatively plausible each contender is, taking into account only “background information.” Background can often be usefully defined as what the analyst knew about the problem when they formulated their hypothesis specifications.

The current investigation is, however, retrospective. In principle, Carrier already knew all the information he currently knows at the time he wrote his specifications. The fictitious nature of the investigation gives him great latitude about what information he selects to assess an initial belief state.

After a lengthy exposition of background information divided into 48 “elements,” Carrier chooses the last element to build his initial belief state. Carrier counts how many items on a list of 22 features of “heroic king” stories (a modified version of the Rank-Raglan index, link) comport with the Gospels.

Carrier estimates that Jesus achieves a score of 20 out of 22. Among all figures who score more than 11 out of 22, Carrier believes at least two-thirds of them are fictional or mythological. Convinced that this cut-off score defines a useful reference class, Carrier proposes an initial belief state of 2:1 odds (the low end of an interval of values that extends up through 15:1) favoring the mythicist hypothesis,

Mythical Jesus: At the origin of Christianity, Jesus was thought to be a celestial deity who communicated with his subjects only through inspiration, such as by dreams, visions and prophecy (past or present). Jesus was believed to have endured ordeals in a supernatural realm involving incarnation, death, burial and resurrection.

over the historicist one,

Historical Jesus: An actual man who was at some point named Jesus acquired followers in life who continued as an identifiable movement after his death. Some of this man’s followers claimed that he had been executed by Jewish or Roman authorities.

There is no obligation for anybody else to choose their initial belief state as Carrier chose his. The Uncertaintist has posted about why fair-minded people might in general disagree about reference class probability estimates (link). Other writers have made specific criticisms of how Carrier scored Jesus and interpreted the result (for example, link).

One serious concern is whether the proposed reference class is relevant to the actual problem being studied. A fair-minded critic might object that not everybody who’s ever been called heroic and kingly “counts equally.” Some heroic kings may be differently situated than Jesus in ways that follow from the undisputed statements we’re assuming as premises:

Partial background: Early in the history of Christianity, some followers began worshiping Jesus as a living god or demigod. At some point, a story of this same Jesus was composed and told within the sacred community. Some or all communities taught that this story was real or both real and allegorical. The story situated Jesus on earth and in history as a divine man, with an earthly family, companions, enemies, deeds, sayings and an earthly depiction of his ordeals.

For example, does it distinguish Jesus among all those who have ever been hailed as heroic kings that the text being scored was consciously composed by or for people who worshiped him as a god, and as part of a dedicated community?

Is it relevant that those writers plausibly knew earlier stories of familiar heroic kings? All of Carrier’s high-scoring legends (Oedipus, Moses, Theseus, Dionysus, Romulus, Perseus, Hercules, Zeus, Bellerophon, Jason, Osiris, Pelops, Asclepius, and Joseph son of Jacob) would have been available to the Gospel authors.

In other words, while there is a defined set of high-scoring kingly heroes, is Jesus a typical member of that set, or can he be distinguished from other members based on background information? If we have such distinguishing background information, why oughtn’t we use it in crafting an initial belief state?

There are plenty of other ways to formalize initial beliefs, and no mandate to use reference classes. The great latitude afforded to Carrier by the fictitious character of a retrospective investigation applies to everyone else, too.

Nevertheless, Carrier reports that he would attain a belief state with at best 2:1 odds against the historicity hypothesis early in this fictitious investigation. That report stands as a valid self-description, regardless of what anybody else thinks of how Carrier arrived at that state. Let’s move on, then, and look at how evidence about Jesus and Christian origins might change the relative plausibilities.

Step 3: assessing how the evidence alters the initial plausibility

Carrier portrays himself throughout the book as arguing a fortiori. That Latin phrase describes arguments that propose two statements which share some feature, but in different “amounts.”

Example: If there’s not enough room left in the garage for a motorcycle, then a car won’t fit, either.

Carrier feels that he’s giving his opponent’s position, the guild’s historicist hypothesis, the best reasonable odds. If the historicist hypothesis fails in the best reasonable circumstances, then it fails in all reasonable circumstances. As for whether Carrier’s generosity helps to persuade the guild, they would have to agree that Carrier is in fact giving their position the best reasonable odds.

There can be no surprise that sometimes the guild won’t agree. The Uncertaintist has written about how one guild member weights Paul’s report of having met the brother of the Lord (Galatians 1:19). He believes that Paul’s report should settle the question of Jesus’ historicity (link). In contrast, Carrier rates it as favoring the historicist position by only 2:1. The guild member probably wants too much weight for Paul’s remark, but Carrier offers nothing conspicuously generous here. The most reasonable weight could easily lie between Carrier’s offer and the guild member’s demand.

Well-justified disagreement also surrounds Carrier’s treatment of the canonical Gospels as potential evidence. Except to assess the Rank-Raglan score, Carrier gives them no weight at all – the canonical Gospels are a wash for him, favoring neither hypothesis.

The Gospels contain the earliest and best evidence for the portion of the historicist hypothesis, “Jesus acquired followers in life.” Paul attests to Peter and others being leaders in a Jesus-focused movement, but offers no hint that they or anybody else had been followers of Jesus in life.

Similarly, it is the Gospels that offer the earliest clear extant support for the portion, “Some of this man’s followers claimed that he had been executed by Jewish or Roman authorities.” Our Paul says nothing about Roman involvement, and his only mention of Jewish involvement (1 Thessalonians 2:15) could easily be an interpolation.

If the historicist hypothesis were true, then we would be thoroughly unsurprised to see these matters addressed in the second-earliest extant Christian writings, given that they hadn’t been conclusively treated before then. Paul’s audience included followers of Peter (1 Corinthians 1:12), and Paul attests to Peter’s missionary activity in Antioch (Galatians 2:11). Thus, there was plausibly knowledge outside of Jerusalem about what Peter taught, and what his credentials were for teaching. That distributed knowledge constrains what the second and third generation Christian authors composed.

If the mythicist hypothesis were true, these portrayals of first generation leaders having once been followers of Jesus would be inventions, plausible enough but not the only plausible scenario for an explanation of how the Jerusalem leaders came to be church leaders. The creative bar is not set high. Paul explains why he was picked to have a commissioning vision from Jesus on a par with Peter’s. God simply chose Paul for that honor in the womb (Galatians 1:15). Maybe that’s how Peter got involved, too.

More inspiring possibilities for invention existed. The earliest canonical gospel repeatedly toys with the idea that the disciples’ teacher is somehow a reincarnated John the Baptist (Mark 6:14-16, 8:28). John himself may have been earthside earlier, as Elijah (1:4-6, 9:12-13). If any story set on Earth is all made-up anyway, as the mythicist hypothesis implies, then here are the seeds of a great story passed over for the familiar, literally pedestrian gospel narrative.

That comparison, how plausible is what we observe given one hypothesis compared with how plausible it is given the other, is the essence of weighting evidence in Bayesian thought. In this case, equipoise is plainly the lower bound of support the Gospels give the historicist hypothesis over the mythticist, not the upper bound.

The Gospels may still be weak evidence favoring historicity and a poor source for many detailed biographical facts about Jesus and his disciples. Nevertheless, they are the principal foundation for the unspecific historicist hypothesis. It is intuitively plausible that they offer some support for it in a simple match-play comparison where the alternative sees the Gospels as wholly invented but places no constraint on what ripping yarns might have been invented.

Next time, we’ll wrap up the investigation with a “sensitivity analysis.” We scrutnize the match’s outcome to identify factors that may most impede guild acquiescence that Carrier’s mythicist hypothesis deserves more plausibility than the generic historicist hypothesis which the guild favors.
Photo credit: Charlie Chaplin, Modern Times, 1936


Leave a comment

Filed under Inference and choice, Knowable historical Jesus

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.