On Friday, September 6, after many contentious exchanges and at least one defense motion for mistrial, Lynch characterized the afternoon recess as a cease-fire. The prosecution would close on Wednesday. That gave Gilford and Garcia extra time to update their files in preparation for the state’s final witnesses and to refine their own. At one point, Carlos told the judge he didn’t know how he’d be able to finish this “friggin’ case.”
The next day, Lynch wrote in his journal that “we’re now over three weeks into the trial. It has been extremely slow-going with the tension between the attorneys and sometimes, between attorneys and me.” They’d already heard more than fifty witnesses and the state wouldn’t rest until two days later. He didn’t recall a case, he noted, with so many close and often novel questions to be decided, and Scott was even tougher than Springsteen because the defense lawyers were “anal-retentive” on questions relating to evidence and the law; plus, they had the benefit of having reviewed the record from Springsteen. “Also,” he wrote, “they have constantly pushed the envelope on cross-examination, often well past proper outer limits.”
In addition to standard and not-so-standard objections, he’d had to deal with “suggestive photo spreads, relevance of codefendants’ acts loosely connected to but not part of the murders; prior consistent statements used to rehabilitate a witness; questions of hearsay and more complex ‘back door’ hearsay; admissibility of statements made as part of standard jail-booking procedures, admissibility of statements casually made to jail guards and others I’ve already lost track of.”
When the time came to present Springsteen’s West Virginia statement, the prosecution called Merrill. Because Rob hadn’t signed anything, Lynch had ordered the state to paraphrase his videotaped confession in some ten to fifteen sentences to be read aloud in no more than five minutes. As in Springsteen, the defendant’s name would be redacted.
But before Merrill could reach the stand, Garcia rose to hotly object to “every line” of the prepared summary. Overruled. Holding a single sheet of paper, Merrill entered the witness box. The first part of his testimony was standard setup: who was there, how the clock-camera worked, how long the interview lasted, at what point the Nagra tape ran out, what they did to assure Mr. Springsteen he was not in custody. Satisfied he’d covered due-process requirements, Smith asked Merrill if Mr. Springsteen had admitted participation in the Yogurt Shop Murders, to which Merrill replied, “Yes, sir.” And when Smith asked what, exactly, the suspect had said, Merrill read this:
He originally said that he did not know about the murders, did not even know they had occurred until he had been interviewed by the police. As the interview continued, he remembered he bought a newspaper and read it in a stolen Pathfinder on the way to San Antonio. During the interview, Robert Springsteen admitted involvement in the murders by telling us that he went into the yogurt shop prior to the robbery and opened the back door so he had a way to get in. Robert Springsteen said he went through the front door, then went to the bathroom. Robert Springsteen said when no one was looking he unlocked and opened the back door. Robert Springsteen said he propped it open by using a folded pack of cigarettes or a rock to keep the door from shutting all the way, saying it wasn’t noticeable unless you were looking right at it. Robert Springsteen said at some point in time he went back that evening. He said he went through the back door. Robert Springsteen said there was a silver .380 automatic handgun used in the yogurt shop. Robert Springsteen said he raped a girl; stated he did not think he ejaculated. He said he shot a girl in the back of the head with the .380 as she was crawling, screaming, and crying. He demonstrated the position that Amy Ayers died, which was the position we found her in after the fire was extinguished. Robert Springsteen talked about hearing a total of five shots, maybe six, but remembered five. And after the robbery, Robert Springsteen said he left the yogurt shop, went to a bridge where he got out of the car and threw up. Then he ended the interview before we were complete.
After a few more quick questions, Robert Smith passed the witness to Garcia, who was intense and intelligent, but as always, Merrill took the heat with equanimity. When asked if in fact the interrogation segment of an interview wasn’t a fact-finding expedition but an accusatory one, the unruffled cop said he would think that was correct, yes. When asked why police found it necessary to question a suspect in a controlled, face-to-face environment, Merrill replied, “Well, you can look at him, read his body, how he’s answering, what he’s saying. Because you’re thinking ahead of him. You’re thinking behind him….”
Garcia: “You think he did it, don’t you?”
Merrill: “Yes, sir.”
On redirect, Smith pushed for the admission of a still photo taken from the Springsteen videotape, even though the two sides had agreed that no pictures from the interrogation would be shown. Garcia strenuously objected but was overruled and the clip of Rob demonstrating the position of Amy Ayers’s body appeared on the screen.
On Wednesday, September 11, the state rested its case and Lynch called for the defense.
—
Garcia’s strategy diverged considerably from Sawyer’s.
“There was no way to win, taking that tack,” Carlos says now. And that if “nobody was prepared to investigate a homicide like this, so what? So they didn’t do a great job with the evidence, so DPS blew it, so what? John Jones did the best he could with what he had. We can talk about that forever, but what’s to gain for our client?”
Instead, the Scott defense attacked what Garcia calls the Paul Johnson Show’s “back-engineering,” in which the detective came up with a scenario he liked and then did whatever it took to pile up a case around it to make it workable enough to sell. To shore up that theory and prove how many other people knew what Mike and Rob had done—and therefore might well have ended up at the top of Johnson’s pyramid of guilt themselves—the two lawyers had submitted requests for some of the 1992 confessions: Saavedra and Cortez, Alex Briones and Shawn Smith, among others. And they dug up media stories, looking for quotes, leaks and rumors that might have fed Mike Scott’s imagination with scraps of information he eventually became convinced were his own. They would call expert witnesses who could explain how memory really worked and how and why false confessions sometimes occurred.
They began by returning to the crime scene. Questioning firefighters, they made quick jumps from that night’s play-by-play to what they were really after: information about visits Hardesty and other cops had made after the arrests and the answers the firefighters had given them about whom they might have told what they’d seen. Most said they’d described certain details—the ice scoop, the number of victims, the position of the bodies, who was burned the worst, what the place looked like—to wives or other firefighters. One said he’d told his wife “about the bodies of three of the victims being stacked in the rear of the business and about Amy Ayers’s location.” Another remembered saying that the “precise” piles of clothing meant that the disrobing must’ve taken a long time. But the court ruled that without firsthand knowledge, these witnesses could not speculate how far their news might have traveled beyond their immediate circle, so the lawyers had to hope jurors would make that leap on their own.
Then the defense brought back Reese Price and Amy Dreiss, the ICBY workers; Jorge Barney, from the Party House; an FBI ballistics expert; and two of Mike’s friends. The next day, they called on John Jones to read Shawn Smith’s confession and on Mike Huckabay to read Alex Briones’s. Manuel Fuentes again acknowledged that he hadn’t reviewed the videotapes, and Gilford himself took the stand to read a summary the defense had put together of the numerous crime-scene details reported by the Statesman. Afterward, outside the jury’s presence, there was a storm of disagreements as both sides and the judge argued about the importance of tips and rumors. Every time the defense asserted its need to further pursue the possibility that Mike Scott had derived many of his so-called memories from information on the street and in the newspapers, the state shouted “irr
elevant.” Exchanges grew spiky. Once Garcia insulted Lynch, Gilford took over and touched on several instances: that two of the girls were “stacked in a certain way” and one was separate from the others; that one girl was choked and strangled; that money was left in the cash register; that the girls were tied up with their own underwear…and other bits of information that were every bit as accurate as Mike Scott’s.
Late one afternoon after the jury had been dismissed, Gilford previewed for Lynch and the prosecution the testimony they expected to hear from their first expert witness, Richard Leo, the next morning. At the time, only criminologists and attorneys knew who Leo was, but in the next few years he would become widely known, having coauthored The Wrong Guys: False Confessions and the Norfolk Four, a book about four U.S. Navy sailors who, under intense police pressure and without any forensic evidence tying them to the crime, confessed one by one, in domino fashion, to a 1997 rape and murder they didn’t commit. Three were tried and convicted; the fourth pleaded guilty; all four went to Virginia state prison until DNA results and a more cogent confession revealed the identity of the real killer-rapist. They were released but not exonerated, which means that to this day they live as legally designated sex offenders and murderers. The New Yorker and PBS’s Frontline coverage would feature Richard Leo.
Explaining why they’d chosen him, Gilford reminded Lynch that when jury panelists were asked if they could imagine confessing to a crime they didn’t commit, the answer was a nearly unanimous no. And if the defense was to prove its main premise—that Mike Scott did exactly that—the jury needed to hear a sound theory based on case studies and an anecdotal explanation of how this counterintuitive thing could happen. Gilford did a good job describing Leo’s expertise, but the judge wasn’t sold. False confession was one thing; what he didn’t like was testimony aimed at drawing conclusions about the effect of coercive questioning on a potential suspect. And as Gilford struggled to make his case, Lynch’s questions became edgier.
Like many judges, Mike Lynch harbored a faint distrust of expert witnesses, considering them hired guns, often academics flown in to lecture the rubes. And while he recognized the importance of specialized knowledge, he was inclined to get touchy when a witness produced anecdotes and called them science. Soft science was as far as he’d go. Thinking people, he maintained, had their own “specialized knowledge” and didn’t need somebody like “the alleged false confession expert, Richard Leo from Berkeley,” to help them draw conclusions. What’s also pretty clear is that, unfortunately for the defense, Lynch couldn’t get past Leo’s close association with Richard Ofshe.
Gilford handled the judge well, having a quicker instinct for deference than Garcia. And the prosecution said they had no problem acknowledging that false confessions existed, but as they’d already made clear, they strongly objected to the need for an expert and would hold out against Leo’s appearance for as long as it took.
Lynch requested case law and more information. The defense provided two hundred pages of Leo’s testimony in another case, as well as a number of articles and examples of case law on false confessions, suggestibility and the effect of police interrogation tactics on decision making. Labeling Leo’s field of expertise junk science, the prosecution presented Lynch with their articles and examples of case law.
The defense also wanted to call a memory specialist. Their first choice, Elizabeth Loftus, who’d written highly respected books on the subject, had to cancel at the last minute, and so they’d turned to Robert Shomer.
Lynch agreed to study the Leo information that night, and would hear arguments for and against admitting the two experts the next day. The exhausted lawyers finally left the courthouse at seven-thirty that night, but the judge headed back to his office.
Taking to his couch, he began reading; at ten, he drove to a nearby twenty-four-hour diner and, over a late meal, kept reading. At home, he reviewed the prosecution’s transcripts and articles. Obviously, Leo was prepared to tell the jury not just how false confessions occurred but also why a person who’d given one didn’t quickly recant, as most of us would expect. Sometimes, he maintained, the suspect remained in a “post-admission narrative” of guilt despite being convinced of his innocence—a theory that would give the defense room to explain why Mike Scott kept going back to be interviewed.
The Supreme Court had set the standard for the admission of expert witnesses in the 1993 decision on Daubert v. Merrell Dow Pharmaceuticals, Inc., in which it established four criteria of reliability: testing, peer review, error rates and acceptability in the relevant scientific community. As gatekeeper of those standards, Lynch struggled. “Ordinarily,” he wrote, “I might let it all come in just to be sure I give the defense every opportunity to develop a full and proper defense.” He also wanted to make sure they got every legitimate dollar’s worth out of the experts they’d hired. But this one rubbed him the wrong way. He didn’t want to give the Berkeley professor a stage on which to parade opinions supporting what Lynch believed was his obvious agenda: to change police interrogation tactics.
The judge lost a lot of sleep trying to figure out “an honest reflection of the state of the law as best as [he] could determine it.” But he still had questions. “Whatever [you] do in these tough situations,” he wrote, “there is always residual doubt, guilt and second guessing.”
The next morning, outside the jury’s presence, he swore in the two experts. Richard Leo went first and, unlike Ofshe, he wore his expertise lightly. On the stand, he identified himself as a criminologist and social psychologist whose specialty was the study of “coercive persuasion or extreme influence in decision-making”; in other words, how an individual makes decisions in a high-pressure environment. He described modern interrogation techniques as psychological methods meant to shape and change a suspect’s perceptions and to convince him that the most rational and sensible decision, given what the interrogator is telling him and what he has come to believe, is to say, “Okay, you’re right, I did it. Now can I go home?”
He went on to explain the basic methods—again, Reid—used in the Scott interrogation. In order to create a context for their strategies, by first accusing the suspect of lying, and then attacking his denials and his belief in the reliability of his memory, Lara and the others had created a situation in which the suspect could lose only by denying his culpability. Not everybody could be broken in this manner, but it did happen. In time, some people came to think they might have committed the crime without remembering the act itself. And then—isolated, alone, with no outside support to urge them to quit talking—they began to speculate, guess, imagine, toss out possibilities, incorporating street rumors as well as information from the press. They might even look to their interrogators for answers.
Garcia then took on memory expert Robert Shomer, who provided a long list of credentials, publications and research studies. He proposed to first discuss how memory worked and then to gauge the validity of theories held by polygraphers and the police. Shomer had seen the videotapes of Scott’s interrogation; he’d also read the transcripts. The defense plan was to present video clips—the one in which Bruce Stevenson gave Scott his memory-as-VCR lesson, the trance induction of Sal Abreo, the visualization exercise—so that Shomer could refute the hypotheticals.
Robert Smith acknowledged the probable reliability of Shomer’s expertise but didn’t see the need to show video clips, since the jury had already seen the tapes. At one point, when Garcia disingenuously asked Smith which evidence they were prepared to admit, Lynch curtly pointed out that he was the one to decide, not the prosecution.
Lynch ruled that Shomer could talk about how memory worked and how persuasion and suggestion might encourage people to remember things that didn’t happen. No video clips. As for Leo, he could talk about decision making under pressure and how people decided to confess to crimes they didn’t commit, but he would not be allowed to discuss “persuaded false confessions” or “post-admission narratives.” Considering Daubert
and his own conscience, the judge had decided that while false confessions did occur, Leo’s theory that police interrogation tactics caused regular folks to believe they were guilty of crimes they hadn’t committed was pure guesswork.
Gilford protested. The entire case against their client was based on an interrogation the defense believed was coercive. To take away the information Leo could provide—with backup support from case studies, wrongful convictions, exonerations and other criminal trials—would gut their defense. When Lynch wouldn’t budge, Gilford threw in the towel. Fine, then they wouldn’t call Leo at all.
As lead counsel, Garcia didn’t have to go along with this, but by then, both men were pretty used up and he didn’t challenge Gilford’s decision. Later, he admitted he should have pulled rank and convinced Dexter to take what they could get. But he didn’t. And so Richard Leo headed back to California that afternoon, as the prosecution had hoped.
Lynch considers Richard Leo a smart man who’s doing important work and thinks that, even hobbled by his decision, his testimony could have helped the defense. But he stands by his ruling. Based on all the literature he could absorb that night, he believes the witness was prepared to venture past the outer limits of good science into conjecture, speculation and “what he thought sometimes happened.”
After the jury was seated, Shomer testified as expected, and afterward the defense called Paul Johnson again to ask him about phone interviews he’d had with Scott in 1998, and the conversations he’d had with Chandra Morgan before the four boys were arrested, in which she told many stories about the night of the murders but never once mentioned Springsteen, Scott or the other two and never said anything about going to the yogurt shop that afternoon. The last defense witness of the day was APD polygrapher Bruce Stevenson, whose revivification theory had been pretty much obliterated, not just by Shomer but also by ordinary common sense.
Who Killed These Girls? Page 30