13 votes, average: 5.00 out of 513 votes, average: 5.00 out of 513 votes, average: 5.00 out of 513 votes, average: 5.00 out of 513 votes, average: 5.00 out of 5 (13 votes, average: 5.00 out of 5)
You need to be a registered member to rate this post.

The Value of Eyewitness Testimony

The first of today’s two-short-posts from new “Boxes” in my New Testament textbook, on a matter of vital importance to anyone interested in knowing about the historical Jesus.


What Do You Think?

Box 13.3  The Value of Eyewitness Testimony


If you want to know about something that happened in the past – whether in a criminal trial or just among your family and friends – you almost always prefer to learn what an eyewitness saw or heard.   And so most of us unreflectively think an eyewitness report is highly reliable.  But is that the case?

Eyewitness testimony has been studied by legal experts and psychologists since the early twentieth century.  The first important case study occurred in 1902.  In a law school in Berlin, a well-known criminologist named von Liszt was delivering a lecture when an argument broke out.   One student stood up and shouted that he wanted to show how the topic was related to Christian ethics.   Another got up and yelled that he would not put up with that.   The first one replied he had been insulted.   A fight ensued and a gun was drawn.  Prof. Liszt tried to separate the two when the gun went off.

The rest of the students were aghast.   But …

To see the rest of this post, you will need to be a blog member.  Membership is easy and cheap.  And all fees go to charity.  So why not??

You need to be logged in to see this part of the content. Please Login to access.

Jesus and Hell
The Difference Between Eschatology and Apocalypticism



  1. Avatar
    JohnKesler  October 31, 2018

    Although inconsistencies in the resurrection accounts are sometimes seen as liabilities, do you think that they instead contribute a ring of authenticity–that Jesus’ tomb really was discovered empty, even if some details are dissimilar?

    • Bart
      Bart  November 2, 2018

      I don’t think so myself. I explain why in my book How Jesus Became God. What they do show it aht there were traditions of an empty tomb some decades later, in my view.

  2. talmoore
    talmoore  October 31, 2018

    Strangely enough, although Bayesian statistics has been terribly abused by the likes of scholars like Richard Carrier, it can actually be useful in figuring the probably of certain recorded historical events. One use, in particular, is in determining the reliability of a particular witness’s testimony. In one classic example, we might want know what is the probably of event B, given that a witness who is accurate a certain percentage A of the time claims to confirm the occurrence of said event.

    So we might ask, what is the probability A that the witness is correct, given that event B actually occurred. In mathematical terms, we’d write P(A|B)
    We would solve this using Bayes’ formula. P(B)*P(A|B) = P(B|A)*P(A)
    That is, the probability the event would occur in general, times the probability the witness testimony is accurate given the event occurred equals the probability the event occurred given the witness testimony is accurate, times probability the witness testimony is accurate.

    Now, of course, extracting these probabilities from a 2,000 year old document is probably bordering on futile, but there are some things we can test. For instance, we can calculate the number of times the gospel accounts match established history versus the number of times they don’t, and we can figure a certain percentage of accuracy — let’s say 75%. Then we can figure the probability of a specific event occurring — let’s say Jesus being crucified on a Friday, which is 1 in 7, or about 14%. So what is the probability that Jesus was crucified on a Friday given that the gospels say he was crucified on a Friday?

    P(B|A) = what we are looking for, the probability Jesus was crucified on a Friday given the testimony that he was.
    P(A|B) = is the probability the gospel account is accurate given the event actually occurred = 75% (conjectured)
    P(B) = random probability it was a Friday = 14%
    P(A) = probability the witness is accurate, calculated using the formula P(A) = P(A|B)*P(B) + P(A|not B)*P(not B) = 0.105 + 0.215 = 32%

    Re-arranging Bayes formula, we get P(B|A) = (0.75 * 0.14) / 0.32 = 32.8%

    • Bart
      Bart  November 2, 2018

      Interesting. How can the conclusion be checked and verified?

      • talmoore
        talmoore  November 2, 2018

        It would require many conjectural assumptions that, alas, are too difficult, if not impossible to verify. The point is just to show how such a probability could be calculated, should the underlying assumptions be verifiable. This is one of the biggest issues with using Bayesian statistics for historical events, establishing the underlying assumptions. It is why Carrier’s conclusions are so flawed.

        As I like to say, the only thing more important than getting the right answers is asking the right questions. Calculating that there’s a 32.8% chance that Jesus died on a Friday might be the correct answer to the wrong question. Like how a logician could reach a sound conclusion from flawed premises.

        • Bart
          Bart  November 4, 2018

          But if a calculation cannot be checked, how do we know if it is reasonably accurate? I.e., why use a method that cannot be supported in any way other than appealing to itself?

          • talmoore
            talmoore  November 5, 2018

            In some cases, you might have a documented set of numbers to work with, such as, for example, the size of armies or, more to the point, the size of the Exodus. I have orthodox Jewish family members who really believe that 3 million Hebrews marched out of Egypt ca.1300 BCE. I point out to them that from the archaeological record — from the number and size of the dwellings we find and the amount of arable land the Egyptians appear to have cultivated — that 3 million people would likely be half of the entire Egyptian population of that time. One would tend to notice such a huge population shift in the archaeological and historical record, but we don’t.

            When records give you numbers, you can often apply statistical methods to those numbers. That’s when Bayes’ formula, for instance, is useful.

          • Avatar
            SidDhartha1953  November 16, 2018

            In general, based on the research, what is the probability that an “eyewitness account” is accurate in its major details 40-70 years after the fact? I think we have one eyewitness to Lincoln’s assassination who was still living in 1965, so that could be the conjectured lower limit.

          • Bart
            Bart  November 18, 2018

            The length of time until an eyewitness’s death had no bearing on the accuracy of his/her testimony.

      • Avatar
        alexc  November 2, 2018

        I don’t think it can be. My understanding is that the calculation is telling us that IF (big if) the gospel accounts are right about history 3 out of 4 times, then we are a little more than twice as likely (32.8%) to see that Jesus died on Friday compared to a random chance(14%).

        It is never going to tell you what day Jesus died on, only how much above or below random chance an event is to occur given our confidence in the gospel accounts. As an expert in the field the best you could do is to refine the value of how accurate the gospel accounts are, and similar to how the Drake equation can tell us about life on other planets, we can make estimates on how likely Friday was the day of crucifixion. But, as i’m sure you are aware, it’s very open to being manipulated to get the answer people want.

        • talmoore
          talmoore  November 6, 2018

          I don’t think there’s actually a 32% chance Jesus was crucified on a Friday. I think it’s actually closer to 100%. My point was to show how Bayesian statistics can be used to calculate the probably of certain purported events.

    • Avatar
      Matt2239  November 2, 2018

      As people talk about him 2000 years after his undignified crucifixion, it is reasonable to conclude that Jesus beat all the odds, however unlikely those odds might be.

  3. Avatar
    Brittonp  October 31, 2018

    As a juror in a recent criminal case, I was amazed by the significant differences in the testimony of two police officers. They were together throughout the event yet you would not have thought it from their reports.

  4. Avatar
    brenmcg  October 31, 2018

    I guess a corollary to this would be that contradictory accounts in the gospels don’t necessarily mean they weren’t written by eyewitnesses?

    • Bart
      Bart  November 2, 2018

      Yes indeed. The quesiton of whether they were written by eyewitnesses has to be decided on other grounds.

      • Avatar
        brenmcg  November 2, 2018

        I think the only way to interpret John 7:22 is that the verse was written by an eyewitness.

        “Moses gave you circumcision (not that it is from Moses, but from the fathers), and you circumcise a man upon the sabbath.”

        • Bart
          Bart  November 4, 2018

          I’m not sure I see why. Lots of sentences from novels are similar, but are fiction.

          • Avatar
            brenmcg  November 4, 2018

            There’s an editorial note in the verse claiming Jesus made a historical error.

            This is inconceivable if its written by someone in 95AD writing about the eternal word of god made flesh.

            Why not exclude the verse or edit “moses” to “the fathers”?

            The only explanation for an editorial note is someone who heard Jesus say it and not wanting to change any word he said.

          • Bart
            Bart  November 5, 2018

            I don’t see why. It could be someone who had an older tradition about what Jesus said (whether or not he actually said it) and wanted to correct it.

          • Avatar
            brenmcg  November 5, 2018

            The author of the gospel is free to exclude or edit any tradition which contradicts his narrative. And Jesus making a historical error contradicts the author’s claim that this Jesus is the word of god made flesh.

            How could someone writing in 95AD include an editorial note which so utterly negates the central theme of his gospel?

          • Bart
            Bart  November 6, 2018

            Happens all the time!!

  5. Avatar
    Matt2239  November 1, 2018

    There are numerous ways to establish the credibility of eyewitnesses. The research that belittles it is just authoritarian propaganda. As for the oral histories upon which the New Testament is based, it’s reasonable to conclude they’re accurate. The reason is because Jesus’s sayings weren’t just said once. They were part of a three-year ministry that his disciples heard so many times it would be difficult to forget them.

    The glaring conflicts like the day on which Jesus was crucified can even be explained. There always exists the possibility that the differing gospels reflect calendar disputes between long-forgotten factions. Even today, Easter in Orthodox Christianity is celebrated on a different day than in the West.

  6. Avatar
    Hormiga  November 1, 2018

    I think I’ve mentioned this before, but the world of UFOlogy provides an illuminating set of examples of how actual eye witnesses report actual, verifiable events.(*) As you might expect, they’re all over the place. The venerable Condon report of fifty years ago has some useful information, and James Oberg has been doing some running analysis of modern cases.

    (*) Meteors, satellites, bright stars and planets, rockets, space junk reentering the atmosphere, etc.

  7. Avatar
    caesar  November 2, 2018

    In the case of the gospels, the only thing I can think of, to at least start to verify the accounts is to see if they correspond to known facts. If they do correspond, such as getting details about a certain location correct, that doesn’t mean the story is true, but I think it would improve the odds. Are there many accounts in the gospels that we can just discount, because they make mistakes about history, geography…?

  8. Avatar
    RonaldTaska  November 3, 2018

    One of my favorite Ehrman books is “Jesus Before the Gospels” which discusses eyewitness accounts in detail and is a very readable book. I highly recommend it. I have no clue how Dr. Ehrman can write so clearly about such matters. He has a “gift” for sure which I envy a great deal. And, of course, with the passage of time, eyewitness accounts change even more. For me, considering this problem, it is surprising not that the Gospels differ so much, but that they are similar at all..

  9. Avatar
    Iskander Robertson  November 7, 2018

    “There always exists the possibility that the differing gospels reflect calendar disputes between long-forgotten factions.”

    i think Dr Ehrman you already addressed this in a post you did, i tried to find the post was i was unable to.

    • Bart
      Bart  November 9, 2018

      I may have done. I’m actually one of those scholars who don’t think calendrical differences explain much in the Gospels (unlike those who like to reconcile contraditions by claiming different authors were using different calendars. No evidence of that so far as I have been able to find)

  10. Avatar
    SidDhartha1953  November 18, 2018

    “The length of time until an eyewitness’s death had no bearing on the accuracy of his/her testimony.”
    I should think it would if there are multiple recountings that can be compared over time, as the research has done with different subjects. The difference here is there is one subject describing the memory multiple times. Wouldn’t that be extremely valuable?

    • Bart
      Bart  November 19, 2018

      Yes, having multiply attesting accounts is much better evidence than, say, length of time after the event.

      • Avatar
        SidDhartha1953  November 23, 2018

        You seem to be suggesting the consistency of a memory over time is irrelevant. I think that would be a surprising position, given the studies you cite that are looking at just that question.

        • Bart
          Bart  November 23, 2018

          No, in fact the consistency of a person’s memory has been shown to depend on things other than accuracy. You should read the literature on it — very interesting. (Quick example: if a week after you experience something you slightly change what you remember happening in retelling it; the next time you retell it you remember it in its retold version rather than its actual version)

  11. Avatar
    Brandan DeLorenzo  December 16, 2018

    Dr. Ehrman,
    You accept that Peter, Paul, Mary, and James had experiences where they were convinced to have seen the resurrected Jesus. Why do you specifically reject “the 12” (which Paul could’ve known about via Peter) “the 500” when Paul says most are still alive, and “all the apostles,” and again, here Peter and John could’ve very well known about this.


    • Bart
      Bart  December 17, 2018

      The visions of Peter, Paul, Mary, and, somewhat, James are well documented. The others are based on what Paul himself says that he has heard — no other verification.

      • Avatar
        Brand3000  December 17, 2018

        But we only have 1 source for Mary’s vision, the last Gos., John, in 95 AD/CE
        What about Cleopas and the other on the Emmaus road?
        James’ only source for a resurrection vision is in Paul, just like the 500, and all the apostles
        Why wouldn’t Paul have good info about the appearance to the 12 from Peter who was part of that group?
        I know you think that group hallucinations are possible, so why not just say maybe there was one, but it was a group hallucination?


        • Bart
          Bart  December 18, 2018

          All the Gospels assume that Mary was the first to hear, and she does see Jesus first outside of John (see Matthew 28:9-10). That James saw Jesus is attested not only in Paul but is assumed in Acts and in later traditions. The traditions about the 12 and the 500 do not have any additional verification (though of course the 11 do).

          • Avatar
            Brand3000  December 18, 2018

            Would you accept an appearance to a group of 11? Many say “the 12” was a general nickname for the apostles, rather than a specific number that were there at any given time. Also, since Paul is our earliest and best source, shouldn’t his writings be given priority?


          • Bart
            Bart  December 19, 2018

            I doubt it; I give reasons in my book How Jesus Became God.

  12. Avatar
    Brand3000  December 28, 2018

    Dr. Ehrman,

    Do you believe in group hallucination and that groups can hallucinate as easily as individuals, and is that how you would explain the resurrection appearances to several people at once in Paul and the Gospels?

    • Bart
      Bart  December 28, 2018

      As easily/frequently, no. But it’s a well attested phenomenon. Again, you’ll have all these questions answered if you’ll read my book.

  13. Avatar
    Charbel98  May 12, 2020


    What is your view on the claim of eyewitness testimony in the gospel of John? Do you think there is probably some direct eyewitness testimony (from the beloved disciple), or you find it unlikely?

    Thank you!

    • Bart
      Bart  May 13, 2020

      I find it unlikely that much of what is reported in John is based on an eyewitness; but I’m open to the idea that teh beloved disciple may have been a real person who wrote some things down that the author had access to.

You must be logged in to post a comment.