In terms of research I often describe myself as a “document geek”, there are a number of reasons I tend to focus on primary or at least secondary documents – one may be that some decades ago I had a very good teacher in a graduate course on historiography. We received an intense and clinical introduction to risks of relying on individuals as sources and his remarks seem to have embedded themselves in my psyche. When I first became involved with research into political assassinations, in particular the murder of President Kennedy, I was fortunate enough to get a complete set of the early Dallas Police hand written statements. That was back before the internet when scoring such materials was a lot tougher than just using Google and a browser.
One of my first lessons in using the statements was how dramatically witness testimony evolved, either expanding or being refocused within days or at most a week. Another point was that what researchers focused on in terms of a particular witness was not at all what the witness had themselves focused on in their statements. Following that I did a lot of work with FBI interviews and reports and quickly learned a) that FBI agents used closed interview techniques rather than open and b) if a subject wanted to talk about something not on their investigation list they closed them down quickly and dug into what their specific interests were – interviews of individuals who knew Jack Ruby are a really good example of that.
In later years, people have asked me why I didn’t really concentrate on follow up interviews. My response is that while I have talked to many of the people whose names come up in regard to Dallas, both witnesses, police, “suspects”, I just don’t view it as being productive research. Not when I’m talking to them four decades after the event, and especially when I was talking to really bright people like Gerry Hemming. I learned I could catch them in misstatements but that only confirmed what I already suspected, which was that I was wasting my time.
More recently, I’ve been spending a lot of time evaluating eyewitness testimony for some new work I’m doing and I’ve turned back to some of the best available professional studies on the subject. My friend Sherry Feister, a Certified Senior Crime Scene Analyst, provides some excellent insight into witness perception – evaluating witnesses to violent crimes, focusing on their observational ability’s during chaotic events such as shootings. Certainly I would recommend her book, Enemy of the Truth, Myths, Forensics and the Kennedy Assassination to anyone who wades into the JFK, MLK or RFK assassinations.
For a broader perspective, I would also recommend reading Elizabeth Loftus work Eyewitness Testimony, it tends to focus on witnesses from a legal perspective, in particular how law enforcement collects witness information and how it is offered and received in court. However the basic research that she offers and illustrates should make all of us stop and seriously reconsider witness information taken after the first two hours or the first two days. In addition, it also calls into question the tendency to rely too heavily on witness detail.
Extensive testing has shown that marked inaccuracies occur in witness reports of the time involved in observed events, the speed of an object or distance of an object – judgement of speed is especially difficult. Witnesses also routinely overestimate the time of their observation and the time of an event – stress or anxiety in the witness magnify the overestimation – duration of an event is perhaps the most common inaccuracy in estimates.
Going beyond problems with the acquisition of information during events, additional and extensive tests show that “forgetting” occurs rapidly and tapers off over time – accuracy suffers with duration but most significantly in the near term. First day data is far superior in terms of accuracy, one week delay can make a major difference. The reason behind that rapid loss of accuracy is part memory dynamics but also the proven fact that new and even erroneous information quickly contaminates observations- whether it is from the news or from simply talking with other individuals.
Post-event information can actually change memory to the extent of firmly embedding new or false information along with original memories, it becomes almost impossible to separate and witnesses seldom accept the difference even when shown proofs. One of the classic tests demonstrates that witness will argue against their own written, immediate statements – claiming weeks and years after the fact that they must have been mistaken in notes made within hours since they remember it definitely a year or a decade later.
Loftus continues with a detailed study of the impact of “retrieval”, providing examples of how different questioning techniques, innocently or intentionally, can actually add concrete details to a witness story. Signs can be inserted or changed in traffic accident reports, Weapons or other evidence related elements can be added or deleted from robbery or even murder testimony.
Lesson learned – if you are serious about the facts, start with first day statements – to anybody – simple statements to the press are fine, written statements are good but it has to be the witnesses’ own words and not as restated by anyone in a media story or police interview. Now obviously we don’t always have this information, but the closer you get to it the closer you are getting to the reality the witness. At least then you reduce the number of issues down to those of witness perception.