Open main menu
TWA Flight 800 investigation
Portal:TWA Flight 800 investigation

NTSB Board Meeting on TWA 800
August 23, 2000

Day 2, part 3 of 4
part1, part2, part4

Jim Hall: We will reconvene this meeting of the NTSB; the board is in the middle of considerations on the accident report on TWA 800. I apologize to the audience for being a few minutes late, myself, and it's always dangerous for the chairman to be late -- you never know if board members may seize the gavel, so I have to be promptly on time particularly for the one to my right (laughter)

John Goglia: You promoted me earlier ... doesn't that give me any right? (laughter)

Jim Hall: Alright, we now are into the final report that's been prepared by staff for presentation on reported witness observations and I ask Dr. Loeb to make the introduction of the presenter for the final report. (final agenda item, I apologize)

Bernardo Loeb: Thank you, Mr. Chairman, Dr. David Mayer will be the presenter and the witness section. Dr. Mayer has worked for nine years with the Safety Board, initially in the office of Research and Engineering, serving first as a statistics and database specialist and later as a safety study manager. I'll take credit for having brought Dr. Mayer to the board as I hired him while I was director of the Office of Research and Engineering at the time. He's worked on accident safety issues in all modes of transportation but has a background in human performance and in 1997, we brought him into the Office of Aviation Safety and he has been serving as the Acting Chief of the Human Performance Division in the Office of Aviation Safety. He holds a private license and has earned a master's and Ph.D. in applied experimental psychology from Rice University. Dr. Mayer?

David Mayer: Thank you. Good afternoon, Mr. Chairman, members of the Board. In the hours that followed the accident, it became clear that many people had witnessed the crash of TWA Flight 800. The media carried several accounts of people who reported seeing flare-like streaks of fire in the sky at about the time of the crash. The FBI began interviewing witnesses on the evening of the accident and, within a week had contacted over 500 witnesses. During this time, safety board investigators reviewed the many witness accounts the FBI was documenting. The board later formed a witness group that conducted about a dozen interviews. The group also studied more than 450 witness accounts prepared by FBI agents. It was clear from the physical evidence the neither a bomb nor a missile strike had caused the explosion aboard Flight 800. In fact, the witness group agreed that the witness accounts, alone, had not proved or refuted any particular cause of the accident. Nonetheless, many witnesses reported seeing a streak of light that looked like a flare or fireworks, that's followed by a huge fireball in the sky. To some people, this suggested that a missile had been used against Flight 800. In fact, the witness reports were the first and only evidence or indication of a missile attack. But they generated public interest. Consequently, it was important for us to determine if the witness accounts were generally consistent with the physical evidence and for us to try determine what the witnesses were observing. In April of 1998, the board received hundreds and hundreds of pages of documents from the FBI. A team of five safety board investigators was assigned to sort through these materials and organize them for study. This was a painstaking and detailed project, and it required many months of work. During this time, a new witness group was formed. This group, which was made up of representatives from Alpha, Boeing, the FAA, the IAM, and TWA, developed a set of procedures and definitions, and began to study the documents.

This required day after day of literally reading and discussing each witness account. The group determined that the FBI documents contained the accounts of 736 witnesses to the accident. To try to understand what a witness would've seen, we should review, first, the break-up sequence of the accident airplane. As the accident airplane was flying near the Long Island Coast, an explosion amid center wing tank had occurred. About three to five seconds later, the nose section departed and began to fall to the water. However, the aft fuselage and wings continued to fly in crippled flight, probably on fire. During the end of the crippled flight, more than 30 seconds after the center wing tank explosion, the outboard portions of the wings failed. Shortly thereafter, the left wing separated from what remained of the ? fuselage of the right wing. This was followed by water contact.

The entire sequence was not an instantaneous explosion. The sequence required almost a full minute to play out. Now, let's consider what might have been visible to someone watching this whole sequence from Long Island. First, the intact airplane would have been visible to some of the witnesses but it would've been very small and not very noticeable. To give you an idea of how small the airplane would've looked to the witnesses, you may not have noticed the 747 on this slide. It's in the upper left corner inside the NTSB seal. It's that horizontal line just above the eagle. If you're sitting in the front row of the theater, this is how large a 747 would appear from about 11 miles away. About three-quarters of the witnesses to the accident were 11 miles away or further. Some were much further. In fact, nearly a hundred witnesses were more than 23 miles away. If you're sitting in the very back row of the theater, this is how large a 747 would appear from about 27 miles away. I should mention, before we continue, that although anyone watching the presentation could get an idea that the point I'm trying to make is that a 747 would appear very small from a long distance, only the people seated in the theater looking at these big screens are getting the exact size that's appropriate for the demonstration.

The center wing tank explosion occurred inside the intact airplane, so it's unlikely that witnesses would have seen this explosion. The nose section broke apart after the airplane's electrical power failed, and it was virtually free of fire or heat damage. In other words, it was not illuminated as it fell, so it was also unlikely that witnesses would have seen this.

During the crippled flight, it is likely that a fuel-fed fire would've been visible to witnesses some distance away. Such a fire would've looked as a small light or streak. When the outboard portions of the wings failed, it opened up the outboard wing fuel tanks. This likely led to a growing fuel-fed fire. To a witness, this could've looked like an explosion followed by a developing fireball. As the aircraft continued to break up, the fireball grew in intensity and may have appeared to split into more than one ball of flame, which would've seemed to disappear from view in the vicinity below the horizon, tree-line or other obstructions.

So, if a witness saw this entire sequence, what we would expect him or her to see, was a streak of light followed by a fireball, which might split into more than one fireball as it fell. Now I'd like to show you a few witness accounts so that we can consider how closely they match the scenario. Bear in mind we wouldn't expect everyone to see the entire sequence.

This witness was in a residence about ten miles from the accident site, looking through a glass door. According to the FBI witness document, he noticed a red flare descending; it was on a slight downward arc from west to east. There was a thin, white smoke trail following a red dot. He watched the red dot for three or four seconds and then he saw a fireball erupt. This witness may have seen some of the last stages of the structural breakup of the airplane.

Another witness was in a restaurant about 15 miles from the accident site. According to the FBI witness documents, she said that she saw a red dot of light that appeared to be a roman candle moving east parallel to the horizon. One or two seconds later, the dot became bigger and began to fall fast. The dot then blew up into a fireball. When a second fireball developed, the two fireballs fell separately. It sounds as if she saw some of the crippled flight and the structural breakup.

Finally, here's the account of a witness on _____ Quouge Beach, about 16 miles from the accident site. According to the FBI witness documents, he said the he saw a flare blossom into a larger flame in the sky. This larger flame came down into the ocean in two columns. Like the last witness, he probably saw some of the crippled flight and the structural breakup.

What these witnesses describe seem to fit our understanding of the accident scenario pretty well. Most of the witnesses were watching the final seconds of the breakup of Flight 800. Let's look at some numbers: of the 736 witnesses, 258 reported seeing a streak of light; 599 witnesses reported seeing a fireball falling from the sky. Of the fireball witnesses, 200 reported seeing it split into two balls of fire.

The witness group concluded that the streak of light may have been the accident airplane in crippled flight. The group further concluded that the reports of the streak of light were generally consistent with the calculated flight path of the accident airplane. Further, the witness group concluded that this fireball was the accident airplane; those who saw it split were observing the final stages of the breakup of that airplane.

Although each witness document was different, countless different words were used to describe the accident -- especially the "streak." When you boil it down, hundreds of witnesses described seeing a streak of light, like a firework or flare, followed by a developing fireball that broke apart as it fell into the water. There is remarkable consistency among the witness accounts: most of them seem to be describing the breakup of the accident airplane, as we understand it from the physical evidence.

Now, it's very common for safety board investigators to interview eyewitnesses, but we do so very carefully. We commonly find that some eyewitness accounts are inconsistent with physical evidence, reported data and with each other. Because of our experience with eyewitness information, with this investigation we expected to find witnesses who disagreed with each other and witnesses who were at odds with the physical evidence. It should not surprise anyone that we found both.

It would be futile to try to fit each witness account into what we know about the breakup sequence, but I want to report that only a relatively small number seem to be at odds with the breakup sequence and, even then, in only a few details.

Let's return to the streak of light. Some people support the suggestion that the streak of light was a missile. Most of the 258 streak of light accounts seem to be describing the flight path of the accident airplane. But there were 56 witnesses who reported that the streak originated at the surface or behind the horizon and/or that it traveled straight up or nearly so. Flight 800 in crippled flight didn't originate at the surface; Flight 800 was never ascending straight up?????

Now, as I mentioned, we weren't surprised to find some accounts that didn't seem to fit. But, there has been speculation that these witnesses were seeing a missile ??? against TWA Flight 800. Consequently, it is important for us to seriously study these reports.

When we considered the number of reasons why people might have reported ???, for example, it's possible that, for some witnesses, as the airplane maneuvered in crippled flight, it appeared to fly nearly straight up. This might have been the streak some witnesses reported as flying straight up?????

The FBI documents, for some reason, or what the witnesses told FBI agents, the FBI did not make any transcripts or recordings of these interviews. Documents are written in the words of the FBI agents who prepared them. Some of the documents contain incomplete information or are vaguely worded. In other words, the documents may not always say what the witness said. Perhaps some of the witnesses who were classified as reporting a "streak" originating at the surface didn't actually say that.

And there's another phenomenon we should consider: memory errors. It's been well-documented that people's memories change over time, and witnesses may incorporate inaccurate information into their memories. Let's talk about this in more detail: the memory is not like a videotape. Psychologist Ira Hymen of Western Washington University has written that people do not simple retrieve their memory and replay the experience. He said that people combine knowledge from various sources with their own personal experience to create memory. In other words, all memories are under construction and these constructed memories change over time.

In the summertime beach community, flares and fireworks are pretty common; we know that flares were set off by boaters after the accident. One of the witnesses told FBI agents that they'd seen a lot of flare activity off the coast in the months before the accident. Many witnesses apparently believed they were watching flares or fireworks. Trying to understand what they had seen -- perhaps even replaying over the event in their minds -- some witnesses may have formed a mental image of the streak, ?? streak which came up from the surface and one that flew straight up. Images like this get inserted into our memories and later recalled as the actual scene. In fact, Hymen believes that almost any activity could cause a witness to think about, imagine or discuss events believed to be the creation of "memory errors."

There are several factors that increase the likelihood of memory errors. These include repeated questioning, information from other witnesses, being questioned by an interviewer who seems knowledgeable about the event, or one who asks leading questions, or afterwards receiving information about the event from multiple sources.

Let's go over this list again. Each questioning provides an opportunity for a witness to imagine the event again, and perhaps inadvertently add more information to the memories. During the investigation, many streak of light witnesses were interviewed and re-interviewed ... some three, four, five times. And many witnesses had access to what other witnesses had said, group interviews or sources such as news accounts. When witnesses are interviewed by someone they believe is knowledgeable about the account, perhaps because they believe the interviewer has access to special information or has talked to other witnesses, witnesses can become more susceptible to memory errors.

FBI agents probably would have seemed knowledgeable to the witnesses and sometimes missile experts were present during the interview. It had to be obvious to many witnesses that these people were interested in whether or not they had seen a missile; this could have influenced the memories of what they saw.

Agents were provided with questions such as, did you see the launch point, and how long did the missile fly. When they ask questions like this, it may have resulted in witnesses reporting surface originating streaks when that isn't exactly what they saw. These questions are very obviously leading but research has shown that much more subtle aspects of question-framing can lead witnesses to describe things that they actually didn't see. Finally, when witnesses are repeatedly exposed to inaccurate information from several sources, there's more likelihood they will incorporate this information into their memories.

Many of the witnesses were on the beach or on a boat with their families or in restaurants with their friends. They witnessed this accident together. There's every reason to believe that they would have discussed it with each other and, if not, with other witnesses several times before being interviewed. Plus, the intense speculation about a missile began almost immediately after the accident. Some witnesses even told FBI agents that they didn't know what they had seen was an airplane crash until they watched the evening news. Witnesses had many opportunities to be exposed to information...

?: Could you please slow down?

David Mayer: The point of what I'm saying is that witnesses had many opportunities to be exposed to information that could have been incorporated into their memories before they were interviewed. No one has perfect recall. Even memories that we are sure of can contain errors. And this is one possible explanation why a relatively small number of witnesses provided information that doesn't seem completely consistent with physical evidence. We understand the breakup sequence very well. The physical evidence tells us a great deal about what happened to the airplane during the minute before it went into the water. There is remarkable consistency among the witness accounts and most of them seem to be describing at least some breakup, some part of the breakup sequence of the accident airplane.

Now I'd like to talk about three specific witnesses. This map depicts the location of TWA Flight 800 as well as these three witnesses. I'm going to talk about a witness who was on the bridge at W. Hampton Beach, NY. He was about 11 miles from Flight 800 at the time of the explosion of the center wing. I'm also going to talk about a witness who was a pilot of the NY international ? helicopter that was operating near Gabresky Airport. He's labeled "helicopter" and is just north of the witness on the bridge. He was about 14 miles from Flight 800 at the time the center wing exploded.

Finally, we're going to talk about a witness who was a passenger in another airplane operating about 6000 feet above, and 3 miles southwest of TWA Flight 800. This witness was aboard US Air Flight 217; that's how his position is labeled.

Let's start with the witness on the bridge. He was about 11 miles from the accident ??? According to FBI documents he said that he saw what appeared to be a chip? fireworks coming off the beach about 4 or 5 houses west of the bridge. It was like a white spark that went up and arched across the sky. A short time later, he observed an orange fireball in the sky to the south of the bridge. This fireball disappeared behind the second house to the west of the public parking area. This account was useful because of its reference to the houses. There are six? reference points to where his observation started, and where it ended.

We obtained an aerial photograph of the bridge, which includes the houses the witnesses used as reference points. Here's the bridge he was standing on. Here's the row of houses he referred to. They're numbered one through five in the parking lot. He first saw the streak about four or five houses west of the parking lot, so we drew a line from his position on the bridge to a point right between those two houses. He said the fireball disappeared below the second house so we drew another line to this house. Then we extended the site lines out to the accident site. Here's what this looked like: this next map shows a portion of Long Island and the accident site. The witness, again, was here on the bridge. At the time of the center wing tank explosion, Flight 800 was here. This line is about where the witness first saw the streak of light. This is about where the accident airplane was when the center wing tank exploded. The second line is about where the witness reported seeing fireballs descend out of his view. This is about at the position of the green zone where the aft fuselage was recovered. In other words, the witness on the bridge first saw the streak of light near where the center wing tank explosion occurred and the fireball disappeared where the airplane impacted the water. He saw a streak and a fireball, moving just like the accident airplane would have moved. His report is fully consistent with the breakup sequence of the accident airplane.

Now let's turn to the witness who was a passenger onboard US Air Flight 217. The airplane was traveling north, he was on the right side of the airplane in seat 5F. According to FBI documents, he said that he observed the blinking lights of another airplane that flew underneath Flight 217. About 15 to 20 seconds later, he observed a light that appeared to be a flare. This flare was visible for about 10 seconds. Then he observed an initial, small explosion in the same area where he last saw the flare. Then a second later, the small explosion turned into a larger one.

Once again, the streak and explosion he described sound much like the witness accounts that I read earlier. It sounds like this witness saw some of the breakup sequence of Flight 800 but some people would suggest that he saw a missile. He said that another airplane passed underneath Flight 217 and then he saw a flare and explosion. We used radar data to study this sequence. First, here's the track of US Air Flight 217 which was at about 22,000 feet. We determined the identify of the airplane that this witness saw pass underneath Fight 217 and used it as a reference point. The other airplane was a US Navy P3 Orion operating at about 20,000 feet. Here's the track of TWA Flight 800. The tracks on this map end at the locations of these three airplanes at the moment of Flight 800's center wing tank explosion.

To help you visualize this better, here is an oversized airplane symbol that represents Flight 217. We found that at the moment the P3 passed out of view, Flight 800 was not yet visible to the witness on Flight 217. That means when the Flight 800 center wing tank exploded, Flight 800 was not yet visible to the witness on 217.

If a missile had been launched, this witness could not have seen it strike Flight 800. We know from the radar data that, at the moment he first reported seeing a flare, Flight 800's center wing tank had already exploded. In other words, the flare that he saw couldn't have been a missile that caused the tank to explode. The tank had already exploded when Flight 800 came into his field of view.

He reported seeing a flare and then a growing explosion. This is consistent with his having seen the latter stages of the breakup of Flight 800.

Finally, let's turn to the helicopter pilot. Many believe that he saw a missile, so let's review what he told the FBI less than 48 hours after the accident. He saw a streak of red light moving very fast, from his right to his left, that moved almost horizontal in a gentle, descending curve. He observed the streak for about one or two seconds and then he saw an explosion and a second explosion, and a large fireball developed and fell to the ocean. Although this account sounds like the final seconds of the structural breakup, some people believe he is describing a missile strike of Flight 800. The Safety Board interviewed him and his co-pilot in 1997. The co-pilot was flying the helicopter and he told us that, after seeing the explosion, the crew immediately flew the helicopter at maximum power to try to render aid. Again, we turn to radar data to study this. The helicopter crew was practicing instrument approaches to Gabresky Airport here. At the moment the TWA Flight 800 center wing tank exploded, Flight 800 was about here. We found the helicopter in the radar data and studied its southbound flight. The helicopter took an unwavering course to the point in the green zone where the aft fuselage was recovered.... the point where the fireball developed in the final seconds of the breakup sequence. Not the spot about 2 1/2 miles west where Flight 800 was when its center wing tank exploded.

The helicopter appears on the radar data about two miles south of Gabresky Airport. About 101 seconds after the center wing tank explosion occurred aboard Flight 800. If the helicopter was traveling at an average speed of 120 knots between Gabresky and the point where it appears on the radar, it would've been en route to the accident scene or about a minute and 58 seconds before the point where it was first seen on radar. This means the helicopter crew began flying to the accident site about 43 seconds after the explosion of the center wing tank. That's the important point. The helicopter crew began flying to the accident site about 43 seconds after the explosion of the center wing tank. In other words, late in the breakup sequence, long after any missile would have been fired. They saw a fireball and they flew out to where the airplane would've been in final seconds of breakup sequence. They didn't see a missile.

Now let's turn to a different topic. As part of our study of witnesses, we conducted a missile visibility test. We did not conduct the test to determine if Flight 800 was struck by a missile. We've known for a long time that it wasn't. Instead, we conducted this test to determine what a missile launch would look like to observers of known distances from a launch point. Military experts, without knowing how far away a shoulder launch missile could be seen, so we conducted a test.

The test was conducted in April, 2000 at Eglin Airforce Base near Ft. Walton Beach, FL. It was conducted in about the same visibility conditions as those on Long Island at the time of the accident. In other words, the position of the sun, the weather and especially the cloud cover was about the same. Observers were positioned at known locations up to 14 miles from the launch site and after each launch site they were asked to describe what they saw. All of the observers saw the missiles and described them as a rapidly rising light.

Using our knowledge of how shoulder launch missiles perform, let's examine what a hypothetical missile attack would've looked like. The rocket motor of the missile would be visible and it would look like a light ascending rapidly for about 8 seconds. Then the motor would burn out and the light would disappear for as much as 7 seconds. After this, a second streak of light, the airplane in crippled flight would become visible. It would be different from the first streak moving slower, then it would develop into a fireball. We carefully reviewed the witness accounts to determine if anyone described a scenario like this, one that began with two sequential streaks of light and concluded with a fireball. We could not find anyone who was describing this scenario.

In summary, we carefully studied and considered all of the witness accounts. The witness reports and the streak of light are consistent with them having observed Flight 800 in crippled flight. They're not consistent with a missile. This concludes my presentation.

Jim Hall: Alright, Dr. Mayer, is what you're telling us is that witnesses can never be relied upon?

David Mayer: No, sir. I'm telling you that we have approached witnesses very carefully. what they tell us may match/may not aspects of eyewitnesses -- it's important to consider with the physical evidence.

Jim Hall: How can we tell if the witness account is accurate?

David Mayer: The only way I know to do so is to compare what witnesses said with the physical evidence and other recorded data. In an accident investigation such as this -- we discussed in great detail yesterday -- we have a wide variety of information: physical evidence as well as radar data from multiple sites, and other recorded information aboard the airplane. It's critical in my mind to compare what witnesses are saying with what we know from other sources and consider them in that light. That said, since I told you witnesses accounts can change over time, its implicit in that in earlier accounts, are probably more accurate than later accounts. So it's important to consider these, especially.

Jim Hall: Is it your conclusion that the witnesses said they saw a streak of light ... didn't see a streak of light at all?

David Mayer: That's an important question. More than 250 witnesses reported the streak so there's no reason to conclude that they didn't see a streak. What we're saying -- That's a lot of people. They probably saw a streak. What we're saying is that small number reported the streak rose from the surface so traveled vertically. Witnesses who reported details such as this weren't entirely consistent with physical evidence but that's only a small number, and even then, it's only one or two details of what they had to say.

Jim Hall: When you come along and incorporate information you get later, into your memory, are you lying or what? Is that what would then be your recollection?

David Mayer: It would certainly be your recollection so in my mind, reporting accurately what's in your mind certainly isn't lying but we have to bear in mind that when eye-witnesses tell us what they remembered, sometimes it doesn't match the physical evidence. We see that frequently in the accident investigations we conduct.

Jim Hall: Is that -- the Board's procedure if I undersigned it, is to do what with eye-witnesses. Can you explain how our approach of eye-witnesses might differ from the FBI's approach?

David Mayer: I'll try to. We consider eye-witness information as one type of factual evidence that we collect in an investigation. I've already listed several other pieces of evidence so I'll focus on witness information. As you well know, the work of the Board is conducted under the party process.

Jim Hall: And who was on this eye-witness list? This is similar to how we do all other aspects of investigations?

David Mayer: That's correct with the exception that some of the witnesses we didn't get to because the FBI initially interviewed. So that is slightly different.

Jim Hall: And that is something, hopefully, with Congress's enactment of the authorization bill that action will be clarified for future accident investigations that both the NTSB and FBI are involved in.

David Mayer: We certainly hope so.

Jim Hall: Go ahead, Dr. Mayer.

David Mayer: When we conduct an interview of an eye-witness, as I said, we do so under group process. Group members would assemble with the eye-witness and each of them, as interviewed, each member would take notes what the witness reported. We would ask questions such as, what did you see, tell us about your experience. Sometimes, if we felt it was prudent, we'd ask for a tape recorder or ask a court reporter to be present at the interview.

Our strategy would be to produce a complete and detailed record of what the eye-witness told us, and I say that because we like to minimize ... so we don't have to interview at a later date. Frequently, we ask eye-witnesses to sign or review a transcript of the interview or interview notes ... I didn't note how interview notes were prepared. I told you that each group member would take notes during the interview. Once the interview was concluded, the group members would reconvene and go over the notes in some detail, and develop from them one set of notes that each member signed as official notes of the interview.

We do our best to avoid asking questions that are presuming knowledge -- in other words, we might ask a witness, did you see the control surfaces of the airplane before it took off? Rather than, did you see that the flaps were set? We know it's very easy to bias a witness and we tried not to. I think that is essentially what we tried to do when we interviewed witnesses.

Jim Hall: And again, I want to acknowledge that normal board procedures were not followed in this investigation, and we are addressing that because that, unfortunately, has added to a lot of the misconception that has been generated around this. Now, if you could show that the airplane did not climb after the nose departed, will that change your analysis?

David Mayer: No sir, although we believe that the airplane did climb after the nose departed. From the work that Dennis Crater and others have done, our analysis is not actually dependent on that.

Jim Hall: Is sound a factor is this analysis you showed us?

David Mayer: Again, we certainly gave sound a great deal of consideration but our analysis is not based on sound so, no, sir, it's not.

Jim Hall: Can you give me an example of someone whose description of the event has changed over time?

David Mayer: Here is an example of a witness who told us -- I don't have a slide for this, I'm afraid -- he told us that he was traveling south on Hazelwood Road when he saw an explosion out over the water. He describes it as a massive fire that descended to the ground and the FBI documents that pertain to this interview say that he did not see any other object in the sky. This interview was conducted 10 days after the accident.

Later in that same day, FBI agents had the opportunity to talk to him again. Remember that I mentioned first he was traveling south on Hazelwood Road and the next interview, later that day, he said he was traveling south on Rogers road when he witnessed the accident. He said he saw a firework or a rocket go up. He said the rocket was orange in color and had fire coming from its trail -- its tail, and he realized that it's not a firework but a rocket. He said it probably came from land not a boat out on the water. It's interesting to me, in a matter of hours, that here is a new detail in his description.

He is not alone, though. I mentioned that the Safety Board interviewed the co-pilot of the national guard helicopter. The helicopter co-pilot was debriefed after the flight on the night of the accident and he reported that he saw a large section of debris engulfed in flames and falling and, that as it was falling a piece of wreckage broke off and he saw it fall in flames, as well. He didn't report seeing anything in the sky before he saw this debris. Certainly he didn't say that he saw a streak of light. A few days later, he was interviewed by the FBI and he reported that he had seen a fireball -- a pyro? or a flare type object in the sky before the flame and debris. We interviewed him in 1997 -- in January, 1997 -- about this discrepancy and he said he didn't remember the streak until the next day. So, that is an example of details being added over time.

Jim Hall: Now, your group did not start work until what, about the middle part of 1998.

David Mayer: That's true.

Jim Hall: And is the first time that the Safety Board had considered what the witnesses had to say?

David Mayer: No, sir, it's not. The FBI, as I mentioned at the presentation, by the end of the first week of the investigation, we had conducted hundreds of interviews and the Safety Board investigators spent most of the first weeks working right in the middle of this investigation, reading as many of the witness reports as he could read while they were being generated. And he kept Mr. Dickinson and other staff advised what the witnesses said.

At the request of the FBI, the Safety Board did not form a witness group at this time. Very early on in the investigation, however, later in the investigation, the board did form a witness group and that witness group interview around a dozen or two eye-witnesses -- folks who were on board the P3 and C130 airplanes that were in the area. And that first witness group also reviewed, I think it was around 450 witness documents that was provided to it by the FBI, and eventually that group ceased operation prior to the public hearing. We didn't take up the topic of eye-witnesses again. So, ...

Jim Hall: Something that was a part of our investigation -- essentially some knowledge of the information from the outset....

David Mayer: Yes, sir, that's correct.

Jim Hall: And, again, maybe I'm drawing a wrong analogy but I assume, if it were a criminal situation with a murder, there are a whole lot of things that are considered. In fact, we've seen recently with DNA analysis, people who were sent to prison as a result of eye-witness accounts, being let out of prison because of the DNA analysis.

So those individuals who've said -- I received a letter from the pilot of the National Guard helicopter and he stated that he felt that the Safety Board ignored the eyewitness information. Is that accurate, Dr. Mayer?

David Mayer: Absolutely not. We just reviewed -- we've considered eyewitness information from the very beginning of the investigation over the last four years.

Jim Hall: Did we interview the pilot and crew of the National Guard helicopter?

David Mayer: We did. We interviewed him and both of, all of the crew members in 1997.

Jim Hall: Did the witness group have access to the result of that interview?

David Mayer: Yes, sir, transcripts were prepared in the interview center. I have the transcripts of his interview here with me. Transcripts were prepared and all of my group members had the opportunity to see them.

Jim Hall: And what did the helicopter pilot tell?

David Mayer: He told us, essentially, details that are very similar to what we said at the night of the accident. He observed a streak in flight for one or two seconds and then he saw the enormous fireball develop.

Jim Hall: I just want that individual to know that he may not agree with the conclusions that are reached as a result of this investigation, but his interviews and reports are not being ignored by the board.

Were your activities restricted in any way, Dr. Mayer, and was the group dictated to in terms of certain tasks that you could or couldn't do or what you could or couldn't say in an interview?

David Mayer: No, sir. There were no restrictions placed on us; we were essentially given a big stack of witness documents and asked to do what we could to study them and to consider whatever kinds of activities were necessary to complete the study of witnesses and that's what we did.

Jim Hall: You know, of course, the FBI had the CIA do an analysis of the eyewitnesses that they made public at the time they suspended their investigation. Is your analysis of the witness accounts dependent on the CIA work?

David Mayer: No, sir, it's not. We conducted an independent project. We didn't do our project in a vacuum; we were aware of their work but our work is not a derivative of theirs.

Jim Hall: And last week, a group put in a full-page ad in the Washington Times including six witness accounts, and what do you think about those accounts?

David Mayer: I actually have the ad with me. I can talk, if you like, very briefly about the witnesses that are mentioned.

Jim Hall: Well, anyone who wants to spend that much money to put a newspaper ad in, I'd sure like you to go through it.

David Mayer: Well, the first witness listed in the advertisement is the witness on the bridge when he saw the accident. This is the witness I described earlier in my presentation, and I told you that his account was consistent with the motion of the airplane and the crippled flight.

The second witness in the ad was the witness who was on Us Air Flight 217 and I explained to you that he couldn't have seen a missile hit TWA Flight 800 because the timing just simple doesn't work out.

The third witness listed in the ad was at the West Hampton Yacht Squadron when he saw the accident. He was out on the deck of the club, there. More than 20 witnesses who saw the accident from that location -- almost all of them reported seeing a fireball and about half of them reported seeing a streak of light, and a few of them there at that location said it appeared to be traveling straight up or nearly straight up. And during the presentation, I offered some explanations why some witnesses might have reported that detail.

The fourth witness listed in the advertisement said he saw, or in the advertisement it says he saw the accident from the deck of his house ... seeing a glowing red object pick up speed and streak out to sea ... then he saw a series of flashes and a fireball according to the ad. However, July 21, 1996 when the interview was interviewed, his FBI document says he saw a red flare descending and makes no mention of some of the other details. In fact, he's the first witness I used as an example in my presentation.

The fifth witness who's mentioned in the advertisement is the pilot of the National Guard helicopter who, obviously, I already described that mistook/studied? the radar data and calculated the time for his departure from the accident site to the rescue field. He said he'd seen a fireball and the breakup sequence of the airplane, not a missile. Actually, we put together a slide of the sixth witness listed in the advertisement -- Slide 37, if I have my numbering right. This witness said that from his boat he saw a red flare come from the horizon, then he said he saw a ball of white light and streaks emerged from this ball. One of the streaks arched up and the other went down. That's what the advertisement says about his recollection. However, in the week of the accident, he told the FBI agents, according to documents we received, he saw a reddish flare ... it was already in the sky when he first noticed it ... and then he saw an orange/red wreckage fall. Again, it appears to be an example of how details have been added overtime.

The final witness who's identified in the advertisement as witness 649 -- you can actually return to what was just on the screen there. Witness 649 described events that certainly do sound like a missile attacking the airplane. This witness came to the FBI's attention very early on in our investigation and early in the days of the FBI investigation. They conducted a line of sight study using the accounts of 11 of these streak-of-light witnesses and Witness 649 was part of the study because he provided some fixed reference points, two flag poles, I believe, that could be used to bound his visual observation. What I mean is, he said that everything he saw occurred between these flag poles. His location is up there marked "school." He was actually exercising at the high school in the area and you see the position of TWA Flight 800 at the point of the center wing tank explosion. It's in that graphic up there. The yellow line that's been drawn shows his line of sight between those two flag poles interests the green zone right at the spot where we found the aft fuselage. So, again, it doesn't appear that this witness was looking in the right location to see where Flight 800 would have been when it would have been struck by a hypothetical missile. So, that's kind of a review of what was in the ad.

Jim Hall: I would like to make it personally clear, at least, for my own personal observation ... my personal view is I'm sure that all these six individuals are fine citizens as well as the individuals who placed the ad in the newspaper. And we certainly -- any time there is an event -- appreciate eyewitnesses coming forward and providing the information that they may have had observed. That information is a part of our investigation. In this situation we've -- and the FBI -- have carefully reviewed all of these witness reports. What's the total of them, Mr. Mayer ... seven hundred...?

David Mayer: 736.

Jim Hall: 736 witness reports over, what, about a two-year period of time of your work?

David Mayer: That's right.

Jim Hall: And I want the individuals to know that, again, all of these observations have been considered as part of this investigation. Member Hammerschmidt?

John Hammerschmidt: Thank you. Dr. Mayer, thank you for that nice presentation. That seems reasonable ... the ideas contained in it, the information contained in it. My only question is, as I was reviewing this section of the report about recorded witness observation, I was reading in the factual portion of the report and got to page 420 where we began a few sections entitled, "Research Regarding Awareness of Witness Memory, " "Perception and Storing Information in Memory" with sub-categories in this section, "Perception of Distance and Motion," "Storing Information in Memory" going on to the next section, "Retrieving Information from Memory" and the next section would be, "Post-Event Information." My question is, we cite a lot of sources. You have made reference in your presentation to Mr. or Dr. Hymen? But we must have two-to three dozens of footnotes in those few pages. My question is, how authoritative are these sources that we have listed?

David Mayer: I think I can answer that they are very authoritative. They represent peer review literature in the field of experimental psychology; they represent the work that has been done. The first citation of the first section that you mention probably represents around one hundred years of research into human memory. The specific area of eyewitness memory has been studied in detail for the last 20 years. Well-known name in that field, of course, is Elizabeth???. She is one of the most active researchers in this field. But she's hardly alone. There have been a wide variety of people who've studied eyewitness memory and recite research from a variety of them, in fact. In selecting the studies to cite in this report, we purposely divided a number of ... cited publications written by a number of researchers and published in a variety of journals.

John Hammerschmidt: Thank you for that answer. I think that just helps complete the picture in those sections.

Jim Hall: Member Goglia?

John Goglia: I don't really have any questions, just a comment. In all of the years that we've done accident investigations, we do use witness statements often, and the errors that are found in them are substantial. That is not that anybody's lying to us; it's just that not everybody has total recall of what they see and, as you said, sometimes other pieces come to your mind and get added into what your memory banks say that you saw. And, it is fortunate, that so many people have felt like we ignored them, but never lose sight of the fact that witness statements, alone, only help to solve the puzzle. Physical evidence is almost always the key and less there is physical evidence to back up the witness evidence, then it becomes very, very just a judgment call and you have really nothing substantial to hang onto.

I was very concerned, early on in this investigation about the witness statements. Those of us who sit on the board and in cocoons, we do not hear what everybody is saying. We do read the letters that are sent in, we do read the newspapers, and we do take that into consideration, and we constantly ask questions of the staff, and we also, independently, look to see if there is anything to substantiate what is said to us, what is told to us. So, we don't ignore anybody. I know, I know, and I talk to other Board members, I know none of us sitting here today ignore anybody. We are trying to put all those pieces until active deliberation. And that this point, I can tell you I am very comforted with what the witness group has said in their action?

Jim Hall: Member Black:

George Black: I would like to thank David for an excellent literature review and report. When it is published, I will encourage people to look at it because it's almost like a textbook and references and cites and, for my 30 years of interviewing eyewitnesses almost on a daily basis from accidents, it's right down the line with all my experiences and virtually everything that's noted in here I have personally observed and it is just an outstanding review, outstanding job of analyzing this database.

David Mayer: Thank you very much, Mr. Black. I'd be remiss if I didn't acknowledge Barry, Denies .... several other people who helped me, Talcum Breeder, as well, who helped me make suggestions and helped me compile this material.

George Black: I'd almost make an observation that one of the things by virtue of my position I got to interview witnesses pretty quickly. In this case, even the FBI was reviewing, interviewing these witnesses, hours, even days, afterwards, and one of the things that we do not know about these witnesses, was what their condition was at the time they made these observations. I noticed that you refer to someone at Yacht Club on an evening during the summer. I suspect, I know what some of their conditions might have been but that is a bit of information we do not have since they were delayed interviews. I would ask you about the value of re-interviewing witnesses after long passages of time and what you think that is worth.

David Mayer: As I mentioned in the presentation, witnesses memories do change all the time. I should even re-phrase that ... it's not that witnesses memories change over time, it is that people's memories change over time. In fact, there was a study conducted after the Challenger accident has been brought up a number of times while we are talking here. There was a study that was conducted, actually the morning after the Challenger exploded, to try to study people's, the jargon is "flashback memory of events."

What this means is ... think for a moment ... everyone has some memory, it's almost a snapshot for them. For many years, people talked about how they remembered precisely where they were and what they were doing when JFK was assassinated. Psychologist Paul ?? believed that the Challenger explosion would be such a "flashback" event for people.

So, on the morning afterwards, I think he pretty much worked all night to develop a questionnaire, then in the morning after the accident, he asked students in class to relate who told them about the event, how they learned about it, how it made them feel in a variety of other details. Then he sat on that information for 2 1/2 years and sought out the same people, and what he learned from the study, is that he has really caused people to think about memory in a different way.

He was able to locate 44 of the students and asked them the same questions and gave them the same questionnaire and what is the surprising finding was that nobody remembered the details in exactly the same way. In fact, a large number of people have changed or were incorrect. That almost every bit of information they have reported about what they were doing and how they learned about the Challenger accident and the distinct trend that he discovered was that people who learned about it from friends or colleagues tended to remember that they heard it from television news accounts when, actually, what they said after the event, was they learned about it from other people, from casual conversation.

This is an example of how people's memories change over time and it really has brought home for me the need to get as much information as you can from a witness as early as you can in an investigation. Now, obviously, unfortunately we can't know every single question to ask. That's why it's imprint to ask people to describe events in as complete detail as they can, as they saw it.

continued on part 4