The Statistics Are Not “Nil”

“Jesus Tomb” Controversy Erupts—Again

Back to “Jesus Tomb” Controversy Erupts—Again

Camil Fuchs, Professor of Statistics, Tel-Aviv University

[The following is a statement circulated by e-mail by Professor Fuchs to fellow Talpiot tomb conference participants.]

Dear All,

Please allow me to break the promise that I made to myself (to stay out of the heated debate after the symposium) and comment on the statistical issue in the final version of the statement initiated by Profs. Meyer and Magness.

The bottom line regarding that issue is this: In my opinion, the statement that “A statistical analysis of the names engraved on the ossuaries leaves no doubt that the probability of the Talpiot tomb belonging to Jesus’ family is virtually nil if the Mariamene named on one of the ossuaries is not Mary Magdalene” is incorrect. I think that by attempting to have a catchy statement, the statement may fall in one of media’s pitfalls similar to the movie’s “one in 600.”

Before going on, let me emphasize: I continue to be convinced that Prof. Feuerverger’s analysis on the cluster of names failed to yield the stated conclusions. After the symposium, in the process of working on a draft for the paper that I intend to submit to the symposium volume, I became more convinced than before that Feurverger’s method of analysis is flawed. I’ll present below a new example which illustrates this point.

But that does not mean that from the point of view of the cluster of names the probability of the Talpiot tomb belonging to Jesus’ family is virtually nil. After all, there is a reasonable consensus (I think) that some of the ossuaries are inscribed (at least as) Yoshua son of Yoseph, Yosah, Maria. Now, if you assume a-priori that the inscriptions in the tomb can contain both names expected from the New Testament as well as other names (and thus the other names don’t disqualify), the Talpiot tomb’s probability of belonging to Jesus’ family cannot be “virtually nil.” At the same time, in my opinion, that probability is far from the one computed under Prof. Feurverger’s provisos.

I shall not bother you with my own calculations, but I add three points:

a) First, regarding two other calculations and the meaning of the results:
(a.1) Based on a completely different set of assumptions, Kilty and Elliot calculate the posterior probability that the Talpiot tomb site is Jesus’ family tomb site as 0.487 (or 48.7%). The posterior probability is the probability of the event given the observed data. I shall not go into their assumptions and calculations, but here is an example which, I hope, illustrates the meaning of the probability.

Imagine for a moment that WE KNOW that the ossuaries in Jesus’ family tomb site should be exactly as those in Talpiot. Furthermore, imagine that WE KNOW that there is also a second tomb site with identical inscriptions. We discover one tomb site. The probability that this is Jesus’ family tomb site is clearly 50% (very close to the 48.7% above). This is how we understand this probability. If we accept Kilty and Elliot’s calculations, there is a toss-up: is this the correct one or isn’t it the correct one.

(a.2) Ingermanson considers the number of persons (not necessarily re-interred in ossuaries) who lived in that period of time and had the following characteristics: (a.2.1) was named Yoshua and his father’s name was Yoseph, (a.2.2) IF he were buried with 2 women and 3 other men with randomly selected names, then (a.2.3) the names of both women were Miriam (in any rendition), (a.2.4) the names of the 3 other names were among the names of Jesus brothers or disciples. Ingermanson allows the 3 other men to have any names among the names of Jesus brothers or disciples (including the possibility of, say, all 3 of them being named Simon). Under these conditions, his calculations yield a value of 11 such individuals if there were 80,000 men in the period, or 8 individuals if there were 60,000 men (more likely). So, you may consider the probability of having one of those configurations as 1 in 8. The number decreases if we don’t allow all of them to have the same name.

b) Second, regarding Feurverger’s method of calculation:

Consider the following configurations of names on 3 ossuaries with male names (excluding Yehuda son of Yeshua): (b.1) the present configuration (i.e. Yoshua son of Yoseph, Yosah, Mattiah), versus the configuration (b.2) Yoseph son of Yochanan, Yosah and Yaakov (James), or another father’s name. The female names are unchanged. My guess is that nobody would seriously consider the (b.2) configuration as a serious candidate for Jesus’ family tomb site. Yet the Feurverger’s calculations show a higher measure of surprisingness in (b.2) than in (b.1)! In my talk, I mentioned an example in the other direction with (b.1) being again the present configuration and (b.2) having Salome instead of Mariamne. In this case the calculations show a lower surprisingness in (b.2) than in (b.1), although I considered Salome to be more expected than Mariamne. But I feel that the first example with the male names illustrates more convincingly that the method does not seem to adhere to the principle that when a configuration of names is considered by all to be closer to the tested hypothesis (that this is Jesus’ family tomb) then the value in the computation has to label the configuration as “more surprising.”

c) The third point is touchier and it refers to the excluded ossuary inscribed “Yehuda son of Yoshua.” In my opinion, unlike what is mentioned in Feurverger’s paper, the issue is not the complexity or the simplicity of the computation of “A son of B son of C,” but the issue is rather basic. Namely, if the hypothesis that Jesus of Nazareth has a son is unacceptable a-priori, then the probability that this is Jesus family’s tomb site is very small (in this case, we have to consider how likely it is to have in the same tomb site a Yehuda being the son of another Yoshua). On the other hand, if the hypothesis that Jesus of Nazareth has a son is a-priori acceptable, the posterior probability is clearly much higher, regardless of the Mariamne issue. But for that, we have to define the provisos differently (and more reasonably).

Related Posts

The Judgment of Solomon, by Raffaello Sanzio (Raphael), c. 1518. Public Domain
May 20
Searching for Solomon

By: John Drummond

May 16
Where Noah Landed?

By: Biblical Archaeology Society Staff

Apr 30
Ancient Samaria and Jerusalem

By: Robin Ngo

Herod Tomb
Mar 24
How Was Jesus’ Tomb Sealed?

By: Megan Sauter

Write a Reply or Comment

Your email address will not be published. Required fields are marked *

Write a Reply or Comment

Your email address will not be published. Required fields are marked *

Send this to a friend