The Washington Post,
 October 24, 2002

Copyright 2002 The Washington Post  

The Washington Post 

October 24, 2002, Thursday, Final Edition


LENGTH: 1367 words

HEADLINE: Reading the Mind of a Killer; Letters and Calls May Yield Clues to Forensic Scientists

BYLINE: Serge F. Kovaleski and Raymond McCaffrey, Washington Post Staff Writers


Two handwritten letters and an unknown number of telephone calls from the serial sniper may contain key bits of information that could allow investigators to draw a more comprehensive analysis of the elusive suspect, experts in forensic analysis said yesterday.

A case that for the first few weeks seemed to have produced very little tangible evidence now has a stock of samples apparently provided by the sniper himself: pages of handwritten notes, paper, plastic bags, telephone calls that could be tracked to points of origin and voice recordings of the killer himself that could be extensively studied by modern technology.

"The longer we can keep him writing, talking and communicating, we can learn more about what makes him tick . . . and any motive investigators can discern from behavioral analysis and field evidence can add depth and a crucial dimension to the manhunt," said Brian H. Levin, a professor of criminal justice at California State University at San Bernardino and a former New York City police officer.

The opportunity to peer into the mind of the serial sniper presented itself on Saturday, when authorities investigating the attack in Ashland, Va., found a belligerent letter they believe was left for them by the gunman. A shorter letter was then discovered by authorities after Monday's fatal shooting of a bus driver in Aspen Hill. Anything left by a suspect in a homicide case has the potential to identify the killer, and investigators initially look for fingerprints and other evidence that could be matched with criminal databases. In this case, an analysis of the first letter has not yielded any usable forensic evidence that would lead to an identification of the sniper, according to two law enforcement sources.

Authorities also were not able to isolate any fingerprints on a shell casing found at the scene of an earlier shooting. The lack of such forensic clues adds credence to the beliefs of many investigators that the sniper is meticulous in his activities and probably wears gloves when firing his weapon or preparing written communications.

Authorities would not comment, however, on whether investigators had isolated any inconclusive clues, such as a partial fingerprint, that might later prove fruitful. "They haven't gotten anything that's usable now, but that doesn't mean they haven't gotten anything," said one official who is briefed regularly on the investigation.

Even if the letters do not yield direct evidence, such as a fingerprint, they do contain much that lends itself to minute analysis by experts. Roger W. Shuy, a forensic linguist whose career has included working on the Unabomber investigation, said that linguistic profiling is a critical tool because criminals can reveal themselves through their use of language.

Authorities "could glean something about the age of a suspect, within a range, depending on whether the clues that are used are age-graded, like teenager talk, for example," he said.

"We can also tell regional variation, meaning where somebody might have grown up," Shuy said. "If he said he dove into the water, instead of dived, we could infer he was from an area in the northern part of the country. The same goes for the use of pail or bucket. Pail is northern, bucket is midland."

Linguistic analysis can also help investigators learn more about an assailant's educational level and even discover attempts to deceive. People can misspell words in an effort to create the impression that they are someone they are not. But when they misspell words that are not commonly misspelled, then it is possible "someone is trying to disguise his education level through writing," Shuy said.

Writing is also a window into a criminal's emotional state. "There are clues to heightened emotion in language, and one of them is "allness" terms, which are words like 'never,' 'always,' 'none,' 'nobody,' et cetera -- words that are not qualified," Shuy said. "This is the language people might use when they are angry, cornered, in emotional distress or excited."

In the case of the Unabomber, Theodore J. Kaczynski, a linguistic analysis of his 35,000-word manifesto painted a portrait of the suspect that turned out to be more accurate than the one investigators were using. The FBI profile depicted a man who was young, perhaps under 25, probably a laborer and possibly from the West Coast.

But Shuy said the linguistic profiling showed that the Unabomber was probably from the Chicago area, about 50 years old, well educated -- possibly with a doctorate, but not in the social sciences or humanities -- and that he had spent part of his life in Northern California. All were closer to the truth.

Harley Stock, a forensic psychologist with training in hostage negotiating, said the sniper's overtures to police have put investigators in a better position to discern what is driving the gunman.

"For example, what is his motivation for contacting them now and his motivation for carrying out these acts," he said. "With spree killers generally, something happens to them in the time before that sets them off."

Stock said that investigators almost certainly are focusing on whether the sniper's demand for money is an actual attempt at extortion or whether it is designed to confuse. "Now that he has taken that part of the country hostage, he can demand things," Stock said. "This could explain why he did not leave these kinds of notes earlier."

Levin, of California State University, said he doubts that money is the impetus for the shootings.

"The currency here is not money, but power, respect and control. At best, money is a secondary motive," he said. "These kinds of sociopaths want to be in control and do not want be crossed."

Before the sniper's voice can be examined by investigators trying to determine such factors as dialect and mental state, audio experts must first analyze the calls that the suspect has made to authorities and that have been taped.

These experts will try to enhance the tapes to make the recording more intelligible and also to pick up other sounds, said Steve Cain, a forensic tape analyst who has worked as an agent for the Secret Service and Internal Revenue Service.

"Let's say there was something going on in the background, say some sounds in a neighborhood or a noisy arcade, they're going to try to bring out as much of that as possible," Cain said.

The aim is to expose "raw data" and to enhance the voice. He said experts could get enough individual voice characteristics to be used as evidence if the sniper is caught and taken to trial. Cain said audio specialists will be looking at the frequencies of the person's voice and the amplitude, among other things. To do that, however, experts need at least 10 or 15 words, and the evidence is of questionable use if the voice is too garbled.

"What makes the human voice unique in my viewpoint is . . . if you get enough words . . . the manner in which they utter those minimum number of words," he said. Part of that "fingerprint," he added, is dialect and pronunciation and any other idiosyncrasies that can be detected.

But if that fails, much can be determined from the written word. In the investigation of the Zodiac killer in San Francisco, Kelly Carroll, a homicide inspector overseeing what continues to be an active and unsolved case, said the letters revealed "megalomania coupled with a need to demonstrate his superiority and a need to demonstrate it in a public forum.

And he did it through the media, Carroll said. "It's kind of a scary thing to say, but Zodiac is probably the prototype for the modern serial killer . . . the media acting as a carrier for this influenza of terror he wanted to spread in the Bay Area."

This need to demonstrate his superiority could be seen in the killer's use of code in letters that "entices the human mind," Carroll said. Eventually, Zodiac sent the code in three parts to three newspapers. It was eventually cracked by a husband and wife who enjoyed doing puzzles.

"They spent a weekend and broke the code," Carroll said.

Research editor Margot Williams contributed to this report.