Saturday, July 18, 2009

The Alleged Unreliability Of Eyewitness Testimony

In our discussion of the resurrection of Christ on the Stand To Reason blog, Jon Curry has been arguing for the unreliability of eyewitness testimony, something critics of Christianity often do. Jon initially didn't include some qualifiers he later added, but instead suggested that he was addressing eyewitness testimony in general. I addressed eyewitness testimony in general when I responded, citing examples like testimony in a court of law and witnesses to a scientific experiment. But Jon would later claim that "Eyewitness testimony, while surprisingly unreliable, is still enough to establish non-outrageous claims."

The evidence he cited against eyewitness testimony in general, later changed to eyewitness testimony of "outrageous claims", was a case cited by the atheist philosopher Arif Ahmed in a 2008 debate with Gary Habermas. You can watch that debate here. Ahmed discusses the case about one minute into the second video clip, if anybody wants to go directly to it. In that case, an attack was staged on a university professor in front of 141 university students. Seven weeks later, the students were given some photographs and asked to identify which one was the attacker. Of those students, 60% identified the wrong person. The person attacked also identified the wrong person.

Here are some portions of what I wrote in response to Jon on the subject:

Eyewitnesses serve as eyewitnesses in a large variety of contexts. Citing an instance of a staged attack, with attributes unfavorable to the preservation of memories, doesn't tell us much about the significance of eyewitness testimony in general. Has Ahmed's illustration convinced scientists to stop relying on eyewitness testimony in their experiments and their gathering of research? Has it convinced courts of law to stop relying on eyewitness testimony? Has it convinced you to stop relying on eyewitness testimony in your everyday life?

You said that you were granting the date and authorship of the New Testament documents for the sake of argument. It follows, then, that we have men like Matthew, Mark, Luke, and John referring to conversations the resurrection witnesses had with the risen Christ, eating meals with Him, etc. They weren't just watching Him carry out an unexpected staged attack at a distance. And it's not as though the witnesses didn't begin trying to remember what had occurred, didn't write anything, and didn't tell anybody else anything about their experiences until "20 years or 30 years or 70 years" had passed. Thus, Luke can refer to the existence of many other accounts in the opening of his gospel.

Richard Bauckham discusses issues like the enhanced memory skills of oral cultures (such as first-century Israel) and the reliability of human memory in his book Jesus And The Eyewitnesses (Grand Rapids, Michigan: Eerdmans, 2006). I quote Bauckham, and discuss this issue further, here. Regarding the alleged gullibility of ancient people, see here....

We have far more than Ahmed's case to go by. That's why I cited Richard Bauckham's discussion of the subject. He cites far more data than what you're discussing.

But even if we limited ourselves to the case cited by Ahmed, what he describes is a case in which 60% of the witnesses to an event misidentified a person involved. But there was agreement on the occurrence of the event. If people who were far from the stage where the event occurred thought they saw different facial features, or some of the witnesses didn't pay much attention to such details, but instead were focused on the more general features of the event, then 60% can be unable to identify the person when asked to do so. But the identity of the person isn't equivalent to the event. As far as I know, from what Ahmed says about the case, nobody denied that the event occurred, thought that the person in question was an animal instead of a human, etc. And I mentioned some of the differences between Ahmed's case and the cases of the resurrection witnesses. You aren't interacting with my comments on that subject either.

You referred to "eyewitness testimony, which is notoriously unreliable". You cited Ahmed's case when asked for evidence. But you haven't given us any reason to think that his case is representative of eyewitness testimony in general. And even his case involves disagreement over some details accompanied by agreement over at least the general outlines of the event.

Furthermore, Ahmed is himself relying on eyewitness testimony when he cites this case. He's citing the research of Robert Buckhout. Not only is he relying on the testimony of Buckhout about what happened in the case in question, but Buckhout in turn was relying on the testimony of others. And you're relying on the testimony of Ahmed, who is even further removed, regarding what was reported about the case in question.

If you want to reject eyewitness testimony as "notoriously unreliable", then you need to make a lot of changes in your life. Stop using arguments that depend on eyewitness testimony, as you've been doing in this thread. Stop trusting eyewitnesses in scientific experiments, in court cases, in your workplace, in your neighborhood, etc.

C.A.J. Coady writes the following about Robert Buckhout, the source Ahmed is relying on:

"The claim that testimony is unreliable amounts to a sweeping rejection of it as a form of evidence. Just how drastic the rejection is supposed to be is never made clear by the critics, but since 'unreliable' means unworthy of being relied upon a remark like Buckhout's constitutes, on a natural reading, a pretty wholesale rejection. But any such rejection is absurd to the point of idiocy. This is exhibited in Buckhout's own article, as we can see if we ask why we should believe any of the results Buckhout reports to us about the experiments on testimony he says he has done and witnessed, or if we ask how Buckhout came by all sorts of information he relies upon in his work and quotes to us as definitively known. Buckhout tells the reader, fully expecting to be believed, that various results were obtained in a classic experiment in the 1930s by Jerome S. Bruner and Leo Postman at Harvard. This is only one of numerous pieces of hearsay that Buckhout produces to support the unreliability thesis. I do not myself object to the hearsay - it is part and parcel of all scientific work, especially in the social sciences - but Buckhout's own reliance is fatal to his unreliability thesis. Testimony cannot be unreliable if its reliability is required to prove that it is unreliable." (Testimony [Oxford University Press, 1992], p. 265)...

Have you consulted any source other than Arif Ahmed's brief description of the case? If not, then all you have to go by are those brief comments. Ahmed tells us that the eyewitnesses were asked to identify the person in question from a set of photographs. The eyewitnesses didn't initiate it. They were asked to do it, and they were given a set of photographs to choose from. That doesn't suggest "some degree of confidence"....

Do you know how easily the attacker could be identified (how unique his appearance was, what he was wearing, the angle at which he approached his victim, how long the victim had to see him, etc.)? Not if all you have is Ahmed's brief description of the case....

I was saying that it [the case cited by Arif Ahmed] doesn't lead us to the conclusion that eyewitness testimony is generally unreliable. I wasn't denying that it would take us closer to that conclusion. But it doesn't take us all the way there.

It's not as though that study is the only data we have. That's why I keep referring you to Bauckham's book and other sources that address a much larger amount of data than the one study you keep referring to.

How can a study involving one type of eyewitness testimony, namely eyewitnesses of one unexpected and relatively unimportant event that may have been brief (I don't know much about it beyond Ahmed's description of it), lead you to a conclusion about eyewitness testimony in general? Not only is it just one case, but, as I pointed out earlier, the witnesses seem to have agreed on the general outlines of what happened. You're assigning far too much significance to the unreliability of 60% of the witnesses on one aspect of one event in one study....

The vast majority of eyewitness testimony is about what you would consider "non-outrageous claims". Why, then, would you cite the case Ahmed referred to, which isn't about "outrageous claims", to counter my claim that eyewitness testimony is generally reliable? If you agreed with me, but wanted to make an exception for "outrageous claims", then why didn't you say so earlier? Why did you make an unqualified reference to eyewitness testimony as "notoriously unreliable", then argue against my dispute of that claim by citing the case discussed by Ahmed? I made it clear that I was referring to eyewitness testimony in general. I cited examples such as eyewitnesses in court cases and in scientific experiments. You had to have known that I was addressing eyewitness testimony in general. But now you tell us that you accept eyewitness testimony in general, but not for "outrageous claims". You've changed your position.

And you give us no justification for the exception of "outrageous claims". What qualifies as "outrageous"? How do you know that eyewitness testimony is unreliable in such cases? If you're defining "outrageous" as "unprecedented", then see my comments above on precedent.

Did you have precedent for the case cited by Ahmed? Did you already believe in one or more scientific studies in which a majority of eyewitnesses were wrong? Even if you did, how did you believe in the first such case? It would have been unprecedented.

I've mentioned specific differences between the case cited by Ahmed and the case of the resurrection. I've linked to an article that discusses eyewitness testimony in more depth, and I've cited Richard Bauckham's book, which takes into account far more data than the one case cited by Ahmed. As I said before, you're assigning far too much significance to the unreliability of 60% of the witnesses on one aspect of one event in one study. You want us to believe that 100% of the resurrection witnesses were wrong, including ones who had lived with Jesus for years, claimed to have seen the risen Jesus more than once, thought they spent enough time with Him to eat meals and have conversations, etc. And you don't want us to believe that they were wrong about one thing related to the events while being right about the general outlines. Rather, you want us to believe that they were wrong about far more than the 60% of witnesses were wrong about in Ahmed's case. It's far easier to be mistaken about the identity of one man you've never met before, who you saw only briefly and unexpectedly, than to be mistaken about seeing Jesus risen from the dead under the conditions described in the New Testament. How does one mistakenly think he experienced such things? If only 60% of the people in Ahmed's case were wrong about such a comparatively forgettable detail, then why think that 100% of the resurrection witnesses would be wrong about not only far more details, but also the general outlines of what happened?

No comments:

Post a Comment