We have noted that a number of scholars have misstated facts in analyzing the statistics of the tomb. We also acknowledge that despite these errors, the Talpiot tomb might not be the Jesus family tomb. Nevertheless, an analysis on the calculations of the names in the tomb should at least be accurate and based on acceptable statistical methods.
See Also: Probability, Statistics, and the Talpiot Tomb
Inside the Numbers of the Talpiot Tomb
By Kevin Kilty
Department of Physics and Engineering
Editor, Bible and Interpretation
Debate over the Talpiot tomb or the Jesus family tomb has subsided, and many scholars consider the matter closed. As much as anything, debate often focused on the filmmaker, producer-director Simcha Jacobovici and his documentary The Lost Tomb Of Jesus. Many scholars have been critical of the History Channel's continued productions of Jacobovici's Naked Archaeologist especially his Exodus Decoded. Jacobovici's lack of scholarly credentials and his antics have angered biblical scholars, and they have asked why the major networks allow pseudo-archaeologists / amateurs such as Jacobovici to star in these documentaries. A number of scholars have insisted that "sensationalism and misrepresentation have become the order of the day. Presenters on these shows often interview archaeologists and then edit their responses to support view points with which they disagree or were never even asked about." This is an accusation frequently made about Jacobovici's productions concerning biblical topics.
We have always believed that if James Tabor and Shimon Gibson were solely featured in The Lost Tomb of Jesus and Jacobovici had remained out of sight in the sound room, the reception of the discovery of the tomb would have been less volatile and the discussions more profitable and learned. An example of a fair depiction that argued that the Talpiot tomb was not likely to be the tomb of Jesus was the National Geographic's In Search of Jesus' Tomb, which was basically ignored by scholars. During the debates over this charged issue, many scholars made claims concerning the tomb that we believe were mistaken and not supported by the evidence.
We have noted that a number of scholars have misstated facts in analyzing the statistics of the tomb. We also acknowledge that despite these errors, the Talpiot tomb might not be the Jesus family tomb. Nevertheless, an analysis on the calculations of the names in the tomb should at least be accurate and based on acceptable statistical methods. In our estimation, the key to calculating the probability of the Talpiot tomb belonging to the family of Jesus of Nazareth is the inscribed ossuary located in the tomb containing the name Yoseh. Despite the erroneous claims by some scholars, Michael Heiser, for example, we have never included the name Mariamne in any statistical analysis of the tomb. We stated a number of times that if we consider the name Yoseh as meaning more than a variant of Joseph, then the probability that this tomb is the Jesus family tomb is 47%. However, if Yoseh is to be regarded as simply Joseph in all circumstances, then the likelihood that this tomb is "The Lost Tomb of Jesus" is only 3%.
We view other scholarly theories on the tomb as doubtful and not supported by the available evidence. Rachel Hachlili has stated that the Talpiot could not be the tomb of Jesus because he is only identified in the New Testament as Jesus of Nazareth. The likely "inscription should be 'Yeshua from Nazareth' or 'Yeshua son of Mariame.'" This argument is implausible. As we have maintained in our other papers:
Out of the 227 inscribed ossuaries listed in Rahmani, "there are only six such ossuaries inscribed with origins or birthplace listed in Judea or its immediate environs . . . place names on ossuaries are so rare among observed inscriptions that Jesus son of Joseph is some twelve times more likely to occur as an inscription than Jesus of Nazareth."
Christopher Rollston has insisted that "even with the small corpus of epigraphic attestations of personal names, even the Talpiot tomb occurrence of "Yeshua bar Yehosep (Jesus son of Joseph) is not unique." Yet when we examined the number of occurrences of Yeshua/Jesus and Yehoseph/Joseph, the distribution of those names do not support Rollston's claims. In 2826 names in all written sources for the Greco-Roman period in Palestine, 330BCE to 200CE, there are only three occurrences of Yeshua/Jesus linked with Yehosep/Joseph, and one of these instances is Joseph, Joshua's brother. Three examples linking Jesus and Joseph cannot be considered typical or common for this period in Palestine.
Jody Magness stated that if the Talpiyot tomb is indeed the tomb of Jesus and his family, we would expect at least some of the ossuary inscriptions to reflect their Galilean origins, by reading, for example, Jesus [son of Joseph] of Nazareth (or Jesus the Nazarene), Mary of Magdala, and so on.
We found very "few ossuaries" that were "inscribed with the names of the deceased person's birthplace or hometown." In fact, we were surprised how few such ossuaries existed. As we mentioned beforehand, the Catalog of Jewish Ossuaries in the Collections of the State of Israel by Rahmani lists 227 inscribed ossuaries and only six such ossuaries are inscribed with origins or birthplace listed in Judea or its immediate environs. Half of the inscriptions in Rahmani's catalog "refer to the deceased and their kinfolk, and there are . . . seventy-three inscriptions that refer to the father of the deceased" as in X son of Y. As we have maintained a number of times, "statistically, an inscription like 'Jesus of Nazareth' would have been exceedingly rare. Jesus son of Joseph matches the archaeological evidence uncovered in the tombs of Jerusalem."
A new book by Darrell L. Bock and Daniel B. Wallace, Dethroning Jesus, is another example replete with errors similar to those of a number of scholars in debunking the Talpiot tomb. In this book, the authors attempt to address a variety of claims that challenge the traditional evangelical views of Jesus. They devote one chapter to Jesus' tomb at Talpiot.
Here, we argue that the authors demonstrate a lack of sophistication in their interpretation of name frequencies and the statistical probability that Talpiot contained the family of Jesus. A more troubling aspect of their critique is they appear to engage in Christian apologetics in defending the resurrection as plausible, but a tomb containing Jesus' family as not. Bock and Wallace have organized their counter arguments to an interpretation of the Talpiot Tomb as the tomb of the Jesus family into three areas: 1) psychological and cultural, 2) DNA evidence, and 3) statistics. Let's look at their cultural counter argument first in regard to its statistical underpinnings.
Throughout their Chapter Six, Bock and Wallace refer to "the hypothesis" as the claim made in the Discovery Channel documentary that this tomb is that of the Jesus family, and we will follow suit. However, they make other related claims at times which we will attempt to distinguish clearly from "the hypothesis." First, Bock and Wallace state that the first claim put forth in the documentary is that
"...the names in one family tomb so closely matched names associated with Jesus of Nazareth, this tomb likely was the family tomb of Jesus. Statistics were trotted out from a statistician in Toronto, Canada, indicating that the chance of this cluster of names being found in one tomb was 1 in 600."
The derisive term "trotted out" does not do the statistical argument justice, nor is Bock's and Wallace's characterization factual. Indeed, the probability of 600 to 1 is, under certain assumptions, just about exactly the probability of finding such an occurrence of names on any tomb in Jerusalem for this period, as we have shown in two of our papers. More to the point, though, this is not what Feuerverger, the Toronto statistician, actually said. We misinterpreted Feuerverger's statement similarly ourselves, but Feuerverger's actual statement is that the tomb has a 599 in 600 chance of being the Jesus family tomb. This is very different from what Bock and Wallace claim.
Bock's and Wallace's final cultural counter argument is that the names involved in the tomb are too common for the hypothesis to be true. As Bock and Wallace point out and then enlist the concurrence of Tal Ilan, Stephen Pfann, and Amos Kloner for support, there are just 16 names among the Jews of first-century Israel that account for 75% of names in use -- each of these names is found among 3% to 9% of the population. Bock and Wallace paraphrase Pfann,
"...these names are not just common, but extremely common."
The term "extremely common" is quite misleading as the statement that these names would be found in almost any Jewish tomb of the time. If one takes, for example, all the occurrences of any form of the name "Jesus" in the Israeli State ossuary collection (Rahmani Catalogue) other than the two in the Talpiot tomb one finds only four other examples in Hebrew/Aramaic and five in Greek, for a total of nine out of 227.
In fact, as Bock and Wallace explain, every expert interviewed agrees that the names are too common to support the hypothesis. However, none of the experts, as far as we can ascertain, is a statistician or mathematician. Moreover, this argument is irrelevant. These are the names of some of the members of Jesus' family. The commonness of the names is not pertinent to any counter argument because the names in the tomb also include a number of the names in Jesus' family. Furthermore, to simply argue that the common occurrence of each name individually equals the common occurrence of a tomb of such names betrays a lack of understanding of probability. Let us provide an analogy.
In a deck of playing cards, each of the kinds of cards, ace to king, occurs with about the same probability (1/13) as do the common names of first-century Palestine. Therefore, in applying the same logic to poker that Bock and Wallace and their experts apply to the Talpiot tomb, a hand of poker containing a straight is common. So is a full house or a royal flush. To carry the analogy further, drawing to an inside straight is not the act of foolishness this common phrase implies, but rather a logical strategy backed up by the frequent occurrence of whatever card one needs to complete the hand. This argument applied to poker is fallacious, and so is the analogous argument applied to the names in the Talpiot Tomb. Commonness of names does not equate to commonness of combinations of names.
We encounter this sort of thinking in every discussion of the Talpiot tomb. There is little realization among scholars that equating likelihood of names in a set (names of Jews in first-century Judea) with likelihood of groups of names (combinations in a tomb) is a fallacy--a misunderstanding for which there is no antidote other than to take a course in probability or to have a statistician explain the fallacy involved.
Finally, among cultural arguments, Bock and Wallace argue that Matthew being in the tomb is problematic. "What is Matthew doing in the family tomb, and why are Jesus' other brothers missing?" We don't know what Matthew is doing in the tomb, and neither do Bock and Wallace, but are Jesus' other brothers missing or not? We would remind the authors that there were four un-inscribed ossuaries. Do they know who was interned in those ossuaries?
Let's proceed into Bock's and Wallace's category number three, the alleged problems with the statistics of the Jesus Tomb. Here they begin arguing for the commonness of names in the tomb once again, but now they recognize that the association of names is important, and that one name, Yoseh, is particularly important. Thus to dispel any significance of the name Yoseh, they produce a chain of logic that begins with the claim that once a name is used in a family it will likely occur again sometimes in a variant form to distinguish it from other, similarly named family members. So, they say,
"....a name like Jose (Yoseh), which is rare on its own, takes a more common presence in a family already containing a Joseph."
The veracity of this claim, which is not at all improbable, only becomes substantiated by calculating a conditional probability of sons being named for fathers--something no one has done yet, especially not Bock and Wallace.
Consequently Bock and Wallace end this chain of logic by concluding that we must accept that Yoseh is actually a variant of Joseph simply because the patriarch of this family is Joseph. Must we accept this? No. Bock and Wallace propose a hypothesis here, and we have data at our disposal to test their hypothesis. In Tal Ilan's compilation there are 138 occurrences of the name Joseph for which we can also find a number of father-son relationships. Of these, 34 involve a father named Joseph, and one involves a grandfather named Joseph. Another 102 combinations involve a son named Joseph, and one involves a grandson named Joseph. In five instances in all combinations of Joseph is the other person named Joseph. In no instance at all is either one of the paired Josephs called Yoseh--despite the belief of Bock and Wallace that we should find Yoseh in the written record commonly when the father is named Joseph.
The common pairs of names between father and son in Tal Ilan's compilation of the name Joseph are the names one would guess on the basis of name frequency alone -- Simon, Jonathan, Yohanan, Lazarus, Judah, Hananiah and so forth. If one were to consider a man named Joseph, then the probability that he would have a son named Joseph also, based on Tal Ilan's compilation, is 4 out of 34. The probability that a man named Joseph would have a father named Joseph is 4 out of 102. The weighted average of these probabilities is 5.9 out of 100. In contrast, the probability based on name frequencies alone is 8.3 out of 100.
According to Bock and Wallace, we should find Joseph combined with Joseph much more commonly than name frequencies indicate, that is more than 8.3 per 100. Yet, our sample, drawn from the only evidence that exists about this claim, produces 5.9 out of 100. Considering the size of the sample involved there is no significant difference between 5.9 and 8.3 out of 100; and so, we conclude there is no evidence among the compilation of the name Joseph in Tal Ilan to support Bock's and Wallace's claim about the frequency of pairs of Josephs. There is no evidence supporting their claim about the use of the name Yoseh in these instances; in fact, Yoseh never appears directly with Joseph. This is extremely important because if Yoseh is a variant of Joseph to be used to differentiate between father and son, we should be able to locate a number of examples. However, there is no example of this combination, Joseph son of Yoseh or Yoseh son of Joseph, anywhere in the written record during the Greco-Roman period in Judea. Thus, Yoseh is never used in the written record to distinguish another member of the same family named Joseph, unless one is willing to admit to such a relationship at Talpiot. Finally, there is no evidence to reject using name frequencies to calculate the probability of name collections on ossuaries.
Finally, the worst part of Bock's and Wallace's statistical counter arguments now follows. "What population should be given to the region?" they ask; then they answer that four million is a reasonable estimate. But nowhere do they identify the "region," nor do they substantiate their population estimate.
Let us make a sensible statement that the people who are buried in Jerusalem died in Jerusalem. In the first century, one did not transport dead bodies enormous distances just to inter them near the seat of their religion. Because of the practice of burying the dead within twenty-four hours, remains stayed near where people died. Thus, the pertinent figure to use in statistics is not the four million of some vague "region," but the 60,000-90,000 that Levine says occupied Jerusalem during the pertinent time.
Thus, because Bock and Wallace's estimate of the possible population to be buried in Jerusalem is 50 times too large, we must divide each of Bock's and Wallace's estimates about men named Jesus by 50 to arrive at reasonable numbers. Consequently, there are only some 30 men named Jesus to consider, rather than the 1560 that Bock and Wallace claim, who has a father named Joseph and mother named Mary--even if we are willing to agree with Bock and Wallace that every man named Joseph, without exception, also had a son named Joseph that used the pet name of Yoseh. Indeed, Bock's and Wallace's entire exercise here illuminates nothing about the family in the tomb, but it does show that one can arrive at any statistical value by choosing the right denominator.
Let us make a brief summary of our points regarding statistical errors made or repeated by Bock and Wallace:
Pointing out, correctly, that names found in the Talpiot Tomb are common, does not mean that the combination is common.
The assertion that a name, like Joseph, will recur repeatedly within any family is not supported by analysis of names on documents or ossuaries. The name Joseph occurs with no significantly greater frequency among descendants of a man named Joseph, than does the name Joseph among the population as a whole.
The additional assertion that a name like Yoseh is needed to avoid confusion when a father is named Joseph is not supported by any evidence. The only known connection between a Yoseh and a Joseph occurs in the Talpiot Tomb, and here it is inferred indirectly through the inscription on another ossuary.
The number of men named Jesus who have a father named Joseph and a mother named Mary is hugely overestimated by Bock's and Wallace's claim to four million people being available for burial in Jerusalem. The actual number is likely to be a factor of 1/50 smaller, and so are all other figures that Bock and Wallace provide in this regard. Moreover, the rare name Yoseh, known for Jesus' younger brother, makes this factor even less than 1/50.
As surprising as Bock and Wallace's failure to consult with statisticians is, more disturbing is their apologetic arguments in defending the gospel accounts of the resurrection. The authors are troubled, "in terms of theology," in any discussion of a spiritual resurrection as opposed to the bodily resurrection. They label the documentary's handling of the resurrection of Jesus as "naïve."
It is clear if there is a spiritual resurrection as one of the scholarly consultants to the program argued, then the likelihood of a tomb with an ossuary inscribed Jesus son of Joseph has some merit. However, we deem the entire discussion of a resurrection, bodily or spiritual, with considerable skepticism. We certainly understand the authors' predicament in "that Christianity claims that its religious faith is rooted in key historical events. As such, the faith would have been and is capable of being falsified if one could show there was no resurrection." They paraphrase Paul in his well-known defense of the event: "if Jesus is not raised and the dead are not raised physically then believers are to be the most pitied of all people for trusting in a false hope" (see I Corinthians 15: 12-18). We assume they agree. The authors surmise if theological consultants had been part of the documentary The Lost Tomb of Jesus, then this problem would have been recognized. What problem? Are we to believe that the resurrection of Jesus, bodily or spiritually, is more plausible than the possible discovery of a Jerusalem tomb that housed Jesus and part of his family? Though we have argued for a probability as low as 3% that the Talpiot tomb could be the Jesus family tomb, we would maintain that the odds of the resurrection of Jesus or any resurrection are much less.
There are numerous errors and contradictions in the resurrection stories that are unacceptable in any historical inquiry. To ignore the discrepancies in the Gospels one must either synchronize these conflicting accounts or maintain one gospel representation over the others based on suspect criteria. Even an argument that Jesus' followers believed that Jesus was raised does not mean the resurrection took place. This is validating a claim that the gospel stories actually substantiate what the Apostles believed was true, based on documents written decades after the episode. We do not think it is profitable that miraculous events be the foundation regarding the validity of the Talpiot tomb.
We would point out that not one of the Gospels provides an eyewitness to the resurrection. Moreover, the Gospel accounts agree that Jesus' family members and the disciples are all surprised that the tomb is empty. Paul's depiction, the first account of Jesus' resurrection in the New Testament, takes place among others, 500 hundred witnesses, and James, and it is not supported by any other Gospel versions (I Corinthians 15: 6-7). There is no appearance of the resurrected Jesus in Mark (16: 1-8). Only in Matthew does an earthquake herald an angel of the Lord descending from heaven to roll back the stone to the entrance to the tomb (Mt 28: 2). While in Luke, two men in dazzling clothes appear before the women in the tomb (Lk 24:1-7). In Luke, Jesus first appears to Cleopas and a companion on the road to Emmaus, though they do not recognize him until that evening (Lk 24: 13-35). We do note many do not recognize the risen Jesus, including Mary Magdalene. Though it was dark, only in John does Mary Magdalene mistake Jesus as the gardener at the tomb (Jn 20:15). Moreover, John has the unique story of that mysterious "other disciple" whom Jesus loved accompanying Peter in a search for Jesus at the empty tomb (Jn 20:2-9). Jesus emerges as a ghost or a spirit in Luke and then vanishes, reappears, and finally is carried up to heaven; all of this occurs on the day that his followers discovered the empty tomb (Lk 24: 30-51). However, in the book of Acts, Jesus appeared before the believers for forty days, and then he was lifted in a cloud taken to heaven (Acts 1: 3-10). The vast majority of scholars are aware of these inconsistencies and many more. Of course, these visions should not be considered unusual; Jesus sightings continue until our own day.
The authors do bring up a legitimate point. Why would the disciples declare Jesus resurrected, lie to their followers, and die for something they knew "they themselves had fabricated"? This is an important point in this discussion. Bock and Wallace have adroitly drawn attention to the fact that "none of the disciples defected, even when faced with suffering and horrible deaths, including stoning and crucifixion."
Yet, this behavior is a matter of speculation. No one knows exactly what transpired after Jesus' death. We suppose that the vast majority of Jesus' followers were never in a position to verify or refute the resurrection stories; thus, a majority of the disciples were never part of a massive deception. They had been told a story of Jesus' resurrection right after the death of their leader, in which the details can never be recovered. His life, teachings, the trial and resurrection were told and retold, changed, edited numerous times, and adapted to a variety of circumstances until some of these versions reached their destination in our surviving Gospels. There need not have been an immense conspiracy among Jesus' followers to sustain the resurrection story. It would only take a few to move the body to another location or even several locations. The final resting place of Jesus' body could have easily been a secret among his family and one or two followers, all who wished to continue the movement based on the teachings of this remarkable individual. Preserving the memory and deeds of a charismatic figure was not unusual in the ancient world.
The resurrection is not a historical event. It emerges when history ends and is not an observable incident. As we've said earlier, scholars are well aware of these inconsistencies and many more. Our basic argument concerning the resurrection is that historians must employ criteria of what is probable or verifiable. We maintain the same concerning the investigation of the statistical issues concerning Talpiot. In both cases, Bock and Wallace did not succeed.
We admire the authors' clear and unambiguous attempt to defend Christian dogma, the validity of the Gospels, and a bodily resurrection, but this is theology, not a scholarly inquiry.
1 For the most incisive critiques of the Tomb see, Pfann, Stephan, "Cracks in the Foundation: The Jesus Family Tomb Story." Also, "Scholar: 'Jesus Tomb' makers mistaken." See Kloner's comments on the documentary as impossible and nonsense in "Jesus, Magdalene & son in Talpiot tomb."
2 Wood, Bryant G., "Debunking The Exodus Decoded." See also, Christopher Heard, "Exodus Decoded."
3 Flesher, Paul, "How Should Archaeology Reach its Public?" See also Cline, Eric, "The Distortion of Archaeology and What We Can Do About It."
4 There was also a fairly full and balanced discussion by six scholars on the tomb site itself, not so much the film, published in Near Eastern Archaeology (69:3-4 2006: 116-137).
5 Heiser, Michael, "When Jesus Tomb Math Doesn't Count."
6 Kilty and Elliott, "Probability, Statistics, and the Talpiot Tomb." and Elliott and Kilty, "Inside the Numbers of the Talpiot Tomb."
7 Elliot and Kilty, "Inside the Numbers" 1.
8 Elliott and Kilty, 2.
9 Elliott and Kilty, 3.
10 Elliott and Kilty, "Probability." 25-6.
11 Rahmani, L. Y., Catalog of Jewish Ossuaries in the Collections of the State of Israel, Israeli Antiquities Authority, Jerusalem, 1994.
12 "Probability." 25.
13 Bock, Darrell L., and Wallace, Daniel B., Dethroning Jesus: Exposing Popular Culture's Quest to Unseat the Biblical Christ, Thomas Nelson, 2007, 193-213.
14 ibid., 196.
15 ibid., 202.
16 Of these nine examples four are unprovenanced, so we know nothing of other names. Two are on the Sukenik "Jesus son of Joseph" ossuary (No. 9) and two come from Sukenik's 1945 Talpiot tomb (Nos. 113 & 114). Sukenik in fact read these two as devotional references to Jesus of Nazareth, along with their cross marks. No.140 is an unprovenanced fragment, that reads "Yeshua" drawn inside what some have seen as a fish—so again it could refer to Jesus of Nazareth. That leaves only four ossuaries with the name Jesus, none of which have names that could fit him or his family: Nos. 121 and 751 have "Yeshua son of Dostas," and "Jesus father of Simonides" respectively. No. 56, unprovenanced, associates him with a woman named Popilia and finally, No. 89 that reads "Johannes of Jesus."
17 Bock and Wallace, 205.
18 ibid., 206.
19 ibid., 206-7.
20 There is tortured logic here where two pages ago they asked "where are the brothers?" and now maintain that this Jose must be a brother since he is the son of the Joseph involved in the tomb, who appears only as the father of the Jesus of the tomb. They have identified a brother.
21 Ilan,Tal, Lexicon of Jewish Names in Late Antiquity, Part 1: Palestine 330BCE-200CE. JCB Mohr, P.P. Box 2040, D-72010 Tubingen, 2002, 150-68. In two instances there is a grandfather-grandson relationship. So the statistics of true father-son relationships is 4 out of 34; and true son to father relationships is 4 out of 102.
22 A better estimate of how often one finds a father-son pair named alike could come from examining all 2826 names in Tal Ilan's compilation, 57. The fraction of 4 out of 34, in the previous note, is a conditional probability that a father named Joseph would have a son also named Joseph.
23 For a careful and sophisticated consideration of all the relevant facts see Camil Fuchs' work on name frequencies related to the James ossuary. "Demography, Literacy and Names Distribution in Ancient Jerusalem" Polish Journal of Biblical Research 4:1 (7), December, 2004: 3-30.
24 Levine, Lee I. Jerusalem: Portrait of the City in the Second Temple Period, Jewish Publication Society, 2002, 343.
25 Bock and Wallace, 207.
26 ibid., 197, 208-209.
27 ibid., 197-8.
28 What this implies is that the Christian communities surrounding Mark and using his Gospel, which becomes fundamental for both Matthew and Luke, held a version of the Christian faith into the 70s CE that lacked a detailed resurrection appearance without any problem at all.
29 Bock and Wallace, 201.
As I observed on 'ye ole blog'-
Actually theology is itself a scholarly inquiry. So though I appreciate Elliott’s good work, I have to take issue with his presumption that theology is un-scholarly whilst historical work is scholarly and the suggestion that if work is mere ‘theology’ it is somehow less viable than historical inquiry.
Truth told, much historical work is un-scholarly and much theology is quite scholarly; similarly, some historical work is scholarly and some theological work is un-scholarly. It all depends on who is doing what and whether or not they know what they are talking about or are mere dilettantes.
#1 - Jim West - 01/22/2010 - 16:07
Mark and Kevin have raised some interesting new points about the name Joseph and its variant Yoseh. I would like to see some clarification on how the weighted probability was done, since it must make some assumption about the average number of sons a man would have.
I think we could reduce the temperature of these discussions if we all agreed on a couple of points:
1) If Jesus was bodily resurrected, then the Talpiot tomb is not his final resting place and there is no probability calculation to be done. Therefore, when doing probability calculations, one should make it clear up front that we are assuming (for purposes of the calculation) that Jesus was NOT bodily raised from the dead. This eliminates all theological objections from the probability calculations and lets us focus on the mathematical/historical issues. In short, we are estimating a conditional probability, where the condition is that Jesus did not rise from the dead.
2) There is no such thing as a full and complete probability calculation for this problem. There is too much fuzziness involved and we lack data for certain critical parts of the problem. However, it's possible to make ESTIMATES of the probability under various simplifying assumptions. One should always make it clear that a given estimate uses a certain set of clearly stated assumptions.
I don't think the question of the Talpiot tomb is resolved yet. Certainly, we can't brush aside the tomb with the statement that "these names are common." A cluster of common names may be uncommon, if the cluster is big enough. In the case at hand, the cluster is not terribly big, but it is big enough to make the problem interesting. The real question is which hypothesis fits the data better: that the tomb belongs to Jesus of Nazareth, or that it does not.
My own estimate of the probability that the tomb belongs to Jesus of Nazareth is around 2% (under the assumption that any man named Joseph could be plausibly nicknamed Yoseh. I chose this assumption after asking the opinion of a few scholars). This estimate is pretty close to Mark and Kevin's 3% estimate under the same assumption.
I have not done the calculation under the assumption that some fraction of men named Joseph would never be called by the nickname Yoseh. But there is certainly no harm in doing that calculation, so long as the assumptions that go into it are clearly stated. (And it never hurts to explain why those assumptions are chosen.)
#2 - Randy Ingermanson - 01/23/2010 - 22:33
Response to Ingermanson, Part I
I agree with Ingermanson that we could and should reduce the "temperature" of these discussions. I also agree with him that assumptions made to enable calculations need to remain firmly in everyone's mind. We (Elliott and Kilty) have maintained that even our largest probability value of (47%) contains enough uncertainty to accommodate many objections, including that the Talpiot Tomb is, well, just another tomb. We agree on all points possible for one person who may see a bodily resurrection as possible, and another person who sees such as impossible.
Also Ingermanson has raised an interesting point with regard to the assumption about family size. We worried about this for a time for the reason that we have no data to estimate such size. However, if one can make a plausible case that families with a patriarch named Joseph name their children in a manner like other families, with child names appearing with the same frequency as the general population, then this issue goes away. If there is no propensity for repeated naming, the family size becomes a non-issue.
Before we tackle this topic, though, let me summarize how this came to have any importance to the debate. Bock and Wallace maintain that a specific name, like "Joseph", is used repeatedly in a family, and as a result we should expect to find a name like "Yoseh" used commonly to distinguish one Joseph from another. It is an argument intended to discount the value of observing "Yoseh" in a tomb with "Jesus son of Joseph". We decided to examine the issue of too many Josephs and at the same time decide if actual data shows that Yoseh is used to distinguish one Joseph from another. We are concerned here with whether or not the name Yoseh is significant in itself--a nickname that one would carry through life, and be inscribed on an ossuary in the way nicknames are inscribed on markers in our own cemeteries.
By way of this second question we note that the name "Yoseh" is found in occurrence with "Joseph" in only one place that we know of. And that is the tomb at Talpiot. How could we not wonder about the significance of it?
#3 - Kevin Kilty - 01/25/2010 - 01:20
Response, Part II
With respect to the question of frequent reuse of a name in a family, we examined Tal Ilan's compilation for the name Joseph. The frequency with which fathers named Joseph name their sons "Joseph" also, is around 1 in 9. This is not significantly different from the frequency of the general population. The frequency with which sons named Joseph have a father named Joseph in the same compilation is 1 in 25 or so. This is about one-half the frequency of Joseph in the general population, but probably the difference is not significant. When we look at order statistics for name frequency we find, in the sample predicated on father being named Joseph, that the name Joseph occurs as frequently as Lazarus and Simon, and occurs more frequently than Jonathan and Yohanan--quite compatible with order statistics from the general population considering such a small sample. The order statistics for the sample predicated by son named Joseph is very compatible with our other sample and with the general population. The order of frequency is Simon, Yohanan, Lazarus, Joseph, Hananiah, and Judah. There is nothing to suggest that Josephs occur more frequently within families with a patriarch named Joseph than within the general population.
How can one explain these observations? One possibility is that families do name children consistently throughout the general population--a sort of nominal hypothesis. The other possibility is that families name sons after their fathers with a preference, but the families are so large that the other sons not named for the father reduce the frequency of "Joseph" down to the level of the general population. In this case, if "p" represents a propensity to name a son after a father (0->1), and "n" is the typical number of sons in a family, then p/n is the probability that fathers named Joseph also have a son named Joseph. If p=1 then n has to be about 9 sons per family in order for us to observe what we have. Clearly this is not credible. Probably a value of n=3 or so is more appropriate, which implies that p is no greater than about 0.33.
If one looks at the other sample, the one predicated on the son having the name Joseph instead, we would expect, on the basis of name frequency alone, that 8% of these sons come from families with patriarch named Joseph. What we observe, in contrast, is only four Josephs among them. The vagary of sampling here has given us an observation of 1 in 25--well below the general population value of 1 in 12. One cannot suggest any credible effect of average family size that explains this discrepancy.
Nowhere do we see any evidence to suggest that families show any preference for naming sons after fathers as Bock and Wallace maintain. Instead, the simplest explanation for our observations is that they appear to follow the naming patterns of the general population--we tentatively accept our nominal hypothesis. Consequently, there appears little motivation for using pet names to distinguish individuals. Most importantly, our examination of Tal Ilan's compilations shows a name like "Yoseh" is not only rare in its own right, as Bock and Wallace admit, but rare even in association with Joseph.
#4 - Kevin Kilty - 01/25/2010 - 17:31
Thanks for the interesting article, and the interesting comments following it. One minor point -- it's "Jacobovici" and not "Jacobvici" as you have here.
Another point on Joseph / Joses often gets missed in the discussions of the tomb. It is almost always taken for granted that we know that Jesus' father was called Joseph and his brother Joses. It is true that one of Jesus' brothers is called Joses in Mark 6.3, but the same brother appears to be called Joseph in the Matthean parallel, Matt. 13.55. So Mark, who does not name or mention Jesus' father, mentions a brother Joses; and Matthew, who names Jesus' father Joseph, also names his brother Joseph. We don't actually have the Joseph / Joses combination in any single text.
#5 - Mark Goodacre - 01/25/2010 - 19:21
Kilty and Elliott have made another useful contribution to the statistical aspects of the Talpiot tomb. In my related article, "The Talpiot Tomb: What are the Odds?", I also endorsed their previous work on this subject.
However, as I pointed out in my article, it seems to me that, as it stands, the central Talpiot tomb questions can not be resolved through statistical arguments. Currently, there are just too many unresolved factual assumptions and too little data.
As Ingermanson states in his comment below, when we think about the Talpiot tomb both statistically and historically it is clear that we can not just dismiss it out of hand.
It is possible, that the Talpiot tomb is the family tomb of Jesus. What is the probability - I don't know. But I am certain that further data gathering and analytic work on this subject should be encouraged and supported.
#6 - Jerry Lutgen - 02/13/2010 - 16:27
Is your 47% based on the James (Yaacov) ossuary included? As you probably are aware the judge in the forgery case involving the ossuary recommended the prosecution drop the case for lack of evidence Q.E.D. the ossuary is probably authentic. www.netzarim.co.il
#7 - Eliyahu Konn - 05/19/2010 - 18:30
The 47% calculation that this tomb is the Jesus family tomb does not include the possibility that the James Ossuary is the missing ossuary from the tomb. The 47% is based on the assumption that the name Yoseh located on an ossuary in the tomb is a unique name and not a variant of Joseph. The name Joseph was a very common name in the first century, and Yoseh was not. However, if Yoseh is to be regarded as simply Joseph in all circumstances, then the likelihood that this tomb is “The Lost Tomb of Jesus” is only 3%. We have not read any compelling arguments that Yoseh should always be considered a variant of Joseph.
#8 - Mark Elliott - 05/19/2010 - 19:20
As a historian, I lack the statistical abilities to test each calculation properly. But I do know that the so-called "commonness" of the Talpiot Tomb names is the result of misleading calculations methods. Still, Eliot and Kilty concluded the probability of 47% that it is Jesus' family tomb. Now what if the names were not THAT common?
Rollston denied any kind of kinship between the Talpiot Tomb names. Such a suggestion leaves too many holes, but for the sake of debate we may state if the names were common as argued, and Jesus' family was just another Jewish contemporary family, then those names could well be of this family.
In other words the "commonness" proves nothing for or against the Talpiot.
When so much is depended on the so called "commonness" of the names on the side of those who deny the Talpiot Tomb, one wonders what would be the conclusions in case the names are NOT common.
#9 - Eldad keynan - 04/30/2011 - 16:14
A loose brick in the construction of the Talpiot Tomb theory is that we don't know the names of the parents of Jesus, nor his brothers, except James.
The earliest sources do not name them. Their names do not appear until years after the fall of Jerusalem when the gospel writers add them to their stories. Paul does not name them. Therefore, all the statistical discussion is about a group of names in a first century tomb and what their frequency rate is in the population. We should pause for a moment to reconsider what evidence we have that these names have anything to do with Jesus as known from pre-70 sources. Perhaps the gospel writers chose those names because they noticed that they were popular names in the culture, seeing them used by Jews of the time in the cities, countryside, in stories and in tombs on bone boxes....
#10 - Phil Arnold - 04/08/2014 - 05:49