Police Changing Tune in Sniper Investigation

Oct. 17, 2002 -- In today's police press conference on the beltway sniper, there were a number of inconsistencies in what investigators are saying today versus what they have been saying in past days. John Miller, who has been covering the story for ABCNEWS, analyzes where the investigation may be going and why police and others may be changing their tune.

On Tuesday and Wednesday, police were very clear in saying good evidence from witnesses indicated they were looking for a cream-colored van with a taillight not working leaving the scene of the shooting Monday night. They said this info came from a witness at the scene. Composite images of two vans were released. Today they said the evidence about the taillight was no longer credible.

ABCNEWS.com: How could they have been so wrong and allowed bad information out there for so long?

John Miller: What they do is separate the witnesses, take initial first accounts and the compare the accounts. They then drill down into their story by questioning them again about their accounts, maybe three or four more times and ask them about how they met another witness. For example, when Linda Franklin was fatally shot at Home Depot in Falls Church, Va., on Monday, shots were fired and bystanders immediately rushed back into the stores. Eyewitnesses found each other and did what was natural — they compared what they saw and essentially contaminated the eyewitness accounts. Police then have to peel back those layers. They must ask if the witness remembered that piece of information of if they heard it from talking to someone else. It's urgent to get information our early when you have anything. … is not the kind of case you can sit on information — if you think at first it's going to be reliable and accurate.

At a press conference on Wednesday, Montgomery County Police Capt. Nancy Demme said one witness told police the shooter used an AK-74 rifle. Police said that weapon could fire the .223-caliber round that has been the sniper's bullet of choice. "The witness firmly believes this is the weapon," Demme said. Yet today, we hear investigators do not want people to focus on this weapon.

ABCNEWS.com: They say the weapon used by the sniper fires ammunition from the .22 caliber family. They were not even as specific as to say it is a .223 slug. Why the inconsistency?

Miller: There is no inconsistency. The witness was adamant then and now. Demme followed up that statement by saying they don't want people to fixate because they can't gauge the accuracy of the information. We can't have it both ways. We can't press them for details and then criticize them for confirming or denying information.

Witnesses to the killings have described seeing an olive-skinned man who may be Hispanic or Middle Eastern leave the scene. Today, reporters were told this description could be "muddying" the investigation. Police have said they don't have enough information to release a composite sketch. Charles Patrick Ewing, a forensic psychologist at the State University of New York at Buffalo, says it's impossible to tell the difference between information and disinformation. "They're talking about how they don't have enough description to compile a composite sketch. I mean, why say anything at all about that?"

ABCNEWS.com: Do you think it's likely investigators are trying to extend their own messages?

Miller: The trouble with the descriptions is that some witnesses didn't agree. So trying to generate a sketch would be a useless document. During Son of Sam, a set of six or seven sketches came out over the course of a year. The profile went from being a guy with long brown hair combed over to a guy with curly hair with a receding hairline to something else entirely. There were half a dozen sketches and ultimately none of them looked like David Berkowitz — but they looked different from each other too. The danger of releasing sketches is that it sends a message to public that this is what they are looking for. Someone may think they know someone who they think could be a suspect, but they may look at a sketch and say 'he doesn't look anything like that' and not report them. This investigative aid can do investigative harm if the source is unreliable. When you see a sketch and see the guy ultimately caught, the sketch should optimally look like a photograph. But that won't be the case here because they don't have a sketch they're confident enough to release.

Dr. Richard Ratner, a criminal psychiatrist in Washington, D.C., said every word spoken by authorities is now "based on some estimate of what you think is going on in his head.

ABCNEWS.com: Are investigators deliberately acting confused or offering pieces of information that may draw the sniper, or snipers, out? Or, are there big problems with interagency squabbling?

Miller: Early on, investigators were speaking frankly and candidly and including their own feelings and thoughts on the case. Since the sniper has continued to kill, investigators have dialed back and switched to asking help from the public and delivering facts without characteristics. They've now take a more "just the facts, ma'am" approach.

Investigators seem concerned about leaked information, but so far, there appears to be little of that except for the information about a tarot card left at the scene of one of the shootings.

ABCNEWS.com: Is this just a smokescreen?Miller: They are concerned about leaking information, and media people trailing witnesses through the streets — to the point that they will go to the media and tell them not to look at them. They have worries. It is very difficult to conduct a serial killer investigation in a fish bowl. You don't want them to see you coming, especially on live TV. However it's almost impossible to conduct in a vacuum because so much depends on facilitating information from the public. It's a double-edge sword and delicate balance for them.