The Money Bag

 

Bible scholars, particularly the Christian ones, are quick to boast about the reliability and fidelity of the Jewish oral tradition to explain away doubts about the period between when Jesus is alleged to have come back from the dead and 30-60 years later when it was first written down by the authors of the Gospels. To be fair, there is a tradition in Judaism where a deliberate, careful effort was made to pass some stories and some information on from master to student. I don’t know the research where the reliability of this tradition has been analyzed. What we have with documents like the Dead Sea Scrolls, I think, are earlier copies of documents that we also have later copies of, so we can compare and check for drift and fidelity in written transmission. But checking the reliability of oral transmission from 2,000 years ago would be a much more difficult matter. No doubt much has been written on it. Here are some reasons to doubt that this method can really do what Christians claim it does. (What follows is a better version of an analogy I’ve used before).

For all of the repetition about the accuracy of the Jewish oral tradition we hear, there are some very basic points about reliability and transmission between people that are often overlooked. The problem is that we often overlook the cumulative effect of having information repeated again and again as it passes through different speakers. A simple example from probability theory can illustrate the point.

Suppose that a bag with a police escort arrives at a courthouse in Los Angeles. We can suppose that is part of the evidence in a trial. A court clerk receives the bag, opens it and finds a large sum of money. The clerk then asks the police who brought it in some questions. It turns out that the bag travelled from New York. Along the way, it was carried by three different police escorts. It changed hands for different legs of the journey. Let’s also suppose that the manifest has been lost so the clerk doesn’t know how much money started the trip in the bag. The clerk does some checking and discovers that there is corruption in the three police departments that had custody so that the general likelihood that a given cop is honest is .8. Let’s stipulate that if a corrupt cop gets custody of the bag, he or she will take some. And if an honest copy gets custody, he or she will deliever it to the next leg of the trip without taking any of it. The clerk wants to answer this question: What is the probability that the money that arrived in my office is the same amount of money that left New York?

The answer is the probability that the first cop will take some multiplied by the probability that the second cop took some multiplied by the probability that the third cop took some, or .8 x .8 x .8. The probability that the amount that arrived in Los Angeles was the same as the amount that left New York is .51. If you add two more cops at the .8 honesty rate it goes down to .32. And that is despite the fact that the majority of cops in each department are honest. If five cops with a honesty rating of .9 escort the money, there is only a 59% chance that all of it will arrive at the destination. If seven cops with a .95 honesty rating excort it, there is only a 66% chance that all of it will arrive without some being stolen. Of you can think of the a system that captures and relays information. It doesn’t take many generations of copies on a copy machine, particularly a poor one, for the text on the original to become unreadable and for the information to be lost partially or completely. What’s important to note here is that even when the links are highly reliable, the cumulative effect of transmission across multiple links quickly diminishes the fidelity of the system. And it doesn’t take many links, even when the links are 95% reliable for the odds to drop off to the point that it is more likely that the information/money did not make it through than the probability that it did. If there were 5 cops relaying the money from departments that were 80% honest, there is a 68% probability that someone stole some along the way.

(These numbers deal with the transmitters. If we add in a multiplier that represents the reliability of humans at reporting miracles--think of Mary telling someone she saw Jesus as being comparable to the first person who filled the bag and handed it to the cops--then the overall probability that Jesus came back from the dead becomes vanishingly small. See The Case Against Christ.)

Matters are made worse by other variables. Suppose the clerk has no independent way to know what was put in the bag in the first place; she was just handed a bag, afterall. Then she doesn’t know if it originally contained drugs, or diamonds, or cash, or bonds. She could ask the cop who handed it to her, or she could check the contents of the bag for some clue. Suppose there is a note inside the bag itself that says “This bag originally contained $10,000.” Then she counts it and finds $10,000. Now can she be assured that all of the original contents of the bag made it to her safely? No, she can’t. Notice that the note is part of the contents of the bag too. For all she knows, there was $100,000 in the bag, or 5 kilos of heroin, and when one of the cops took $90,out, or replaced the heroin with $10,000, she wrote the note and stuck it in there. Using the contents of the bag itself to determine that fidelity of the system that transmitted the bag is circular and completely unhelpful. What she needs is some independent (trustworthy!) source to corroborate the origination and transmission of the bag. If she put the money into the bag in New York, and then flew to Los Angeles with it, keeping her eyes and hands on it all the way, then she could be more assured (although a person’s honesty with themselves and even their witnessing an event are issues in many circumstances).

The point of the extended analogy should be clear. We are told by a book that has been transmitted to us across 2,000 years and countless unknown people in between that there were some important religious events that transpired in 30-35 C.E. Between those alleged events themselves and the first recording of those events into a system with relatively high fidelity (writing), there were 30-60 years. And during those several decades we do not know how many times the story was repeated or how many people it passed through before it got the authors and they wrote it down.
We have some semi-independent means of secondary corroboration. We have other historical grounds to think that the oral transmission tradition in Judaism at the time was fairly reliable. Part of our evidence is using written sources to check the error rate of stories that were written vs. relayed orally in different eras of history where we have both streams of information. But as far as we know, the stories about Jesus were spreading far and wide among the early Christians in the first two centuries. And while there may be some transmitters who have a higher fidelity than others, we’re not sure who or how many sources the authors of the Gospel stories consulted. There may be a stream of information running through the Jewish oral tradition that is more reliable, but there can be no question that people will talk, and when normal people talk and repeat stories, we know that they embellish, omit, alter, and improve either deliberately or unknowingly. We can see that the story of Jesus’ resurrection varies greatly among the Gospels. And we also know that a number of non-cannonized sources that gave even more contrary accounts were deliberately excluded. So it is difficult to accept some of the exaggerated claims about the reliability of the verbal transmission of the stories.

(Via Atheism: Proving The Negative)
Matt McCormick
Fri, 04 Sep 2009 04:17:00 GMT

Related articles by Zemanta

• The Case for Christ (reasonabledissent.com)
• Pseudoreality and Archaeology. Virtual Reality and Qumran (drjimsthinkingshop.com)