by Edward Humes
When Somers was through with Rosenlieb, Gary Pohlson decided not to attack her credibility for changing her story and lying to the police. If he went after her hard, Pohlson feared she might blurt out something even worse, perhaps about the attorney Teri Bjorn. At that stage in the trial, the judge had not yet ruled on whether Bjorn could testify about her conversation with Pat about Sandy’s disappearance, money and power of attorney, and Rosenlieb had not mentioned her during her direct testimony. Pohlson wanted to keep it that way. Kate left the stand blinking in surprise, astonished that she hadn’t been roasted alive.
Pohlson took the same hands-off approach with Marie Gates, even though her testimony ended up doing much more damage to Pat than Kate Rosenlieb’s had. Gates repeated her story about running into a distraught and weeping Sandy one day in mid-June, two weeks or so before the disappearance (which Somers continually characterized as a “few days” before). Marie recalled for the jury how she heard Sandy announce her plans to divorce her husband because she had made a “terrible mistake” in marriage. And she recalled how Pat grew coldly angry after Sandy disappeared and ordered both his mother and Marie to stop talking about Sandy and to “Keep your dang mouths shut” because it would all “blow over.” Laura and the defense team listened to this, frustrated, as they knew from interviewing Lillian Dunn that Pat’s mother could refute this last claim. But Lillian was unavailable, thanks to Pat’s brother Mike spiriting her off to Orange County.
Instead of grilling Gates, Pohlson tried to counter her emotional testimony by recalling to the stand Kevin Knutson, who said Sandy’s last words to him were instructions to set up a trust that would give Pat greater control over their finances, so that, as she explained, “If anything happened to me, Pat would be taken care of.” These were not the words of a woman about to divorce a husband, Knutson suggested.
But Pohlson did not go after Marie Gates directly, a tactical decision that left Laura uneasy, though she conceded it appeared to make sense at the time. Pointing out each of Marie’s inconsistencies would have been a time-consuming exercise that could easily confuse the jury as well as introduce extraneous allegations that Marie had spouted in the past. It should have been clear from Knutson’s testimony that Marie was just plain wrong, Pohlson reasoned, and Laura tended to agree. Why risk alienating the jury by attacking the grandmotherly woman? Even if the jury believed her, and Kate Rosenlieb, for that matter, Pohlson felt little damage was done. So what if the Dunns had arguments? So what if Sandy threatened him with divorce two weeks earlier, a threat she never acted on? If that proved murder, then half of the men in America could be on trial here, Pohlson said. Pat wasn’t on trial for being a bad husband, nor was it a crime to be happy your spouse died, as Somers seemed to be suggesting. The prosecutor had to link Pat Dunn to the crime of murder itself, and for that, he needed Jerry Coble. And Gary Pohlson just knew he’d burn the chair out from under that witness. Eager to get to the meat of the case, he let Marie Gates go after just a few minutes of cross-examination, seemingly unconcerned by her testimony.
But if Pohlson’s logic was so sound, Laura found herself thinking, why, then, did John Somers look so remarkably content? Sure, the prosecutor had just established at least some evidence of a motive, and he had a timetable for the killings, thanks to changing details in the stories of a neighbor and housecleaner. But if he was worried about an approaching courtroom land mine named Jerry Lee Coble, Laura couldn’t see it. Instead, as he approached the heart of his case—something he had to know the defense was well prepared to attack—he looked positively delighted with the course of the trial. And that worried Laura.
The reason for Somers’ contentment would soon become clear. The prosecutor had indeed reached the heart of this case. But to everyone’s surprise, he did not use the words of Jerry Coble to get there.
He used the words of Pat Dunn.
3
JOHN SOMERS HAD DECIDED TO MAKE PAT DUNN, NOT Jerry Coble, his best witness. Pat didn’t have to take the stand for this. In fact, the strategy worked better with the defendant sitting mute, looking befuddled and angry on the sidelines—a posture Pat seemed to be adopting with increasing regularity as the trial progressed.
No, the prosecutor didn’t need Pat on the stand to use his own words against him. Instead, Somers pulled out sheriff’s reports, transcripts and a host of witnesses to chronicle every utterance Pat had made about Sandy’s disappearance over the course of months. Every minor detail, every niggling discrepancy, every inconsistency—what Somers called “the lies”—became fodder for a full-scale assault on Pat’s credibility and character.
The prosecutor knew, as did Gary Pohlson, that there were no smoking guns in this case, no hard evidence that fingered Pat as a murderer—just Coble. But he also knew that the law allowed juries to decide that a defendant’s lies about important facts in a case could prove “consciousness of guilt,” the theory being that innocent people tell one consistent and truthful story, while the guilty weave a web of falsehoods to cover their tracks. Keeping track of the lies can be difficult, the theory goes, which leads to inconsistencies. A classic prosecution axiom is that the truth never changes, only lies do—at least when the words are the defendant’s.
To demonstrate this principle, Somers brought in a procession of witnesses who purported to prove that Pat could not keep the story of Sandy’s disappearance straight. The fact that Pat was so cooperative, talking to the detectives many times for many hours—and drinking copiously before and sometimes during these interrogations—ended up being used against him in this regard. Every inconsistency, real or imagined, arising during these many conversations became evidence of guilt in Somers’ carefully crafted presentation. There were no big whoppers to be found in Pat’s many statements about Sandy’s disappearance, no inconsistency that could be proved an outright lie in the traditional sense. So Somers went for quantity, knowing that while one or two minor deviations in Pat’s story would seem like niggling details, ten or twelve might have a far different impact on the jury. Somers even produced a huge chart of the “lies” for the jury’s benefit, a virtual road map of deceit he said proved Pat Dunn could not tell the truth about the simplest matters.
First in this lineup came Valley Braddick, the dispatcher at the sheriff’s department who took Pat’s missing-persons report late in the afternoon following Sandy’s disappearance. On the tape of the call, jurors heard Pat giving his first rendition of what happened to Sandy. He was somewhat vague on details, saying at various junctures that he woke up to find his wife missing at nine, nine-thirty, or ten o’clock at night. It was clear he hadn’t been keeping track of exact times, and thus expressed uncertainty. Then, he said, at nine thirty or ten he started driving around looking for his wife and asking everyone he encountered if they had seen her. He returned home around midnight, then spotted the black dog in the yard and Sandy’s keys on the counter—both of which he thought had previously been missing. He immediately resumed his search in the wake of this discovery, first around the house, then back out on the streets. He returned around two, went to sleep for a couple hours, then recommenced the search at four thirty.
Point by point, this first account matched the statements he would next make to sheriff’s detectives. But Somers, anxious to build in as many contradictions in Pat’s story as he could find, misquoted the tape, telling jurors in argument and on his big chart that Pat informed the dispatcher he “got up and drove around from nine to twelve.” This was incorrect—Pat had given a range of possible times, from nine to ten for his discovery that Sandy was gone. Even the prosecutor, who was so critical of Pat’s inconsistencies and so certain that they revealed a guilty state of mind, apparently could not keep all the times straight, either.
Pat Dunn explained this sequence of events next to Kate Rosenlieb, who then recounted it to Detective Dusty Kline. Kline, in turn, recorded Rosenlieb’s story in an official report. According to this thirdhand account, Sandy went to bed at 6 P.M. on July 2. Ro
senlieb made a mistake here, one of greater magnitude than any Pat could be accused of making, as the true date in question was June 30. Somers carefully avoided mentioning the July 2 error—he even told the jury that Rosenlieb had said it was June 30 all along. Otherwise, this account exactly mirrors what Pat told the dispatcher—in fact, he told the same consistent story to Rosenlieb twice, on July 3 and July 4.
Yet Somers still argued that Pat was being deceiftul, on account of his supposedly telling Rosenlieb that he did not want to report Sandy missing because, as Kate put it, “he didn’t want anyone to know.” The obvious conclusion to be made here—since Pat already had filed a missing-persons report days before this conversation with Kate—was that Rosenlieb was the incorrect or inconsistent one, not Pat. Instead, the prosecutor listed this statement about not wanting anyone to know as another of Pat’s “lies.” The sheriff’s department already did know, the prosecutor said, yet Pat concealed this fact from Kate—a lie. Why a murderer would lie to a friend to make himself look more suspicious was never explained, yet that was Somers’ curious argument to the jury.
After his talks with Kate Rosenlieb, Pat next gave an account of the disappearance to Detective Kline directly. Again, what Kline heard mirrored all the other renditions, except Kline wrote in his report that Pat returned home from his second bout of searching at 4 A.M. and slept until 6—a two-hour discrepancy from the statement to Valley Braddick, which Somers seized upon as another “lie.” This is the first genuine inconsistency in the prosecutor’s litany. Kline, however, did not write notes or tape-record as he interviewed Pat, instead crafting his report a full day later. The defense argued that it was just as likely the detective made the mistake in chronology as it was Pat, given the poor record of accuracy in other sheriff’s reports in the Dunn case.
But Somers wanted to show a pattern of deceit and he kept at it, with no detail too small to earn a place on his chart of lies. He brought in the engineer, Roger McIntosh, who, seven weeks after the fact, had told detectives about a conversation in which Pat supposedly gave yet another time for Sandy’s last, fatal walk. McIntosh put the time at two-thirty or three in the morning, instead of the nine to ten o’clock range that Pat had told everyone else. Somers pronounced this “the biggest and most major conflict in his statements,” though the prosecutor expressed no such concern when McIntosh could not remember other details of this long-ago conversation—the one incriminating fact was enough. Laura, for her part, reckoned that the engineer simply confused Sandy’s normal waking and walking hour with the time of her disappearance.
Another witness, Judith Paola Penney, the late Pat Paola’s niece, was summoned to provide proof of Pat’s supposed lies, but she ended up being the one with inconsistent recollections. She testified that Pat had told her of noticing Sandy missing at one o’clock in the morning—which would be a discrepancy—except that Penney had previously told detectives that Pat said three o’clock. On the stand, she admitted that her memory on the subject was vague. Yet despite this, Somers put only one set of times on his chart where it referred to Penney, making her appear far more reliable in her recollections than was warranted.
Pat gave another statement to Detectives Soliz and Kline on July 9, nine days after Sandy vanished. According to Soliz’s report on this interview, Pat reiterated the same basic story and times he had given in all of his other statements to police. Every detail was consistent except for one small point: Soliz reported Pat arrived home at 10:30 P.M. to find the dog in the yard and the car keys on the counter, when earlier statements placed this at midnight. Neither detective seemed to notice this discrepancy at the time, although they immediately pounced on and demanded explanation for any inconsistencies arising in other areas of questioning. As was his practice, Soliz used no tape recorder and took few notes during this interview with Pat, and he waited three weeks to document it. The defense pointed out this delay, arguing that it was just as likely that Soliz had made this minor mistake as Pat.
Finally, Somers turned to the marathon interrogation of July 23, when a team of three investigators questioned Pat nonstop all night while others searched the house for clues to link him to the murder. Again, no notes and no tape recordings were made, while the three detectives took turns interrogating Pat. Their quarry was apparently so consistent in his statements that night that the detectives did not bother to submit separate reports detailing what he said to each of them. Instead, Soliz summarized for his colleagues by simply writing: “Dunn kept giving the same answers or he would say he had already answered those questions and for us to ask new questions.”
However, out of that fifteen hours of search and interrogation, Somers still managed to find one alleged inconsistency from the police report: At one point, Somers stated, Pat told Detective Soliz that Sandy went to bed at 3:30 or 4:00 in the afternoon, though previously he had always said she went to bed at 5:30 or 6:00—another glaring example, Somers argued, of how “it was getting tough to keep his story straight.”
The problem with this point was that it, too, was as questionable as most of the others. On the witness stand, Somers asked Soliz about the July 23 interrogation, wanting to know when Pat said Sandy went to bed that last night. Detective Soliz immediately responded, “To the best of my recollection, it was about 5:30 or 6:00 P.M.”—completely consistent with Pat’s other statements, and contrary to the prosecutor’s big chart. Only after a perturbed Somers suggested that Soliz take another look at his report—this one was written seven days after the interrogation—did the detective change his testimony to say Pat had said 3:30 or 4:00.
The jury had no way of knowing it, but even after looking at his report, Soliz didn’t quote it quite accurately in his testimony. The report did not say 3:30 or 4:00, as he claimed. It says Sandy went to bed at 15:30 or 16:00, using the twenty-four-hour “military” time equivalent of 3:30 or 4:00 P.M. It is also one simple typographical error away from 5:30 or 6:00—the times Soliz initially remembered when questioned on the matter.
In short, a majority of the discrepancies cited by Somers can be disproved, explained or dismissed as the recollections of witnesses who were at least as inconsistent as Pat Dunn was made out to be. And even if there were a few small inconsistencies here and there, the defense argued, they meant little. Did saying four in the morning instead of two in the morning make Pat a murderer? Or just a confused, stressed-out husband who had a habit of drowning his grief in alcohol?
Still, Somers knew that the simple visual presentation of all of those entries on his chart would leave an indelible impression on the jury. Never mind that many of them were reaching, highly questionable or just plain wrong. When the arguing was over and the lawyers sat down, the prosecutor knew that big chart of “lies” would still be there, helping the jury decide what to think about Pat Dunn.
• • •
Still avoiding Jerry Lee Coble, Somers next turned to another important element of his case: attempting to prove only Pat had the opportunity to kill Sandy. As in all of Somers’ cases, there were several components to this theory that fit together like puzzle pieces into a neat whole. First, Pat told detectives that Sandy slept in the nude, and her body was found naked rather than in her normal walking clothes. From this, Somers asked jurors to conclude she must have been killed in bed. Next, there was the fact that a pair of Sandy’s glasses and her jewelry remained in the house—a grave discrepancy in Pat Dunn’s story, Somers argued, because Sandy always wore her glasses and jewelry on her walks. Finally, the prosecutor accused Pat of telling an unintentionally revealing lie, one that only a man who committed the murder would tell: Pat, according to Somers, had claimed that Sandy was wearing a blue jogging jacket when she disappeared—yet that jacket later turned up in the Dunn home.
“The defendant told the police and others that she was wearing that garment when she left the house. . . . The fact of the matter is that that jacket was still at the house because Sandy Dunn didn’t leave the house wearing it. She left the house wr
apped in sheets and a blanket or something similar—because she was dead.”
Again, the prosecutor had formulated a persuasive argument, weaving together a pastiche of small bits of circumstantial evidence that, alone, proved little, but that together might amount to something—were they true.
But the fact that Sandy’s body was found nude really proved very little. The bodies of people kidnapped, molested and murdered by strangers outside the home have on occasion turned up unclothed, even in Kern County. Dana Butler, for instance, was found only partially clad, with the rest of her clothing never recovered—but no one suggested this meant she had to have been killed while home in bed. And there were signs that Sandy could have been attacked by some sort of sexual predator, as her body had been mutilated in the rectal-genital area. For the killer to have done that to her, she would have had to be stripped. (Rape could not be proven or ruled out, due to the advanced state of decomposition.)
As for the jogging jacket that Somers found so devastating, the truth was that Pat Dunn never said he knew Sandy had worn it that night. Contrary to the prosecutor’s argument, Pat initially said he had no way of knowing what Sandy was wearing, because he was asleep when she left. It’s right on the tape-recorded call of Pat reporting Sandy missing: “I don’t know, because she has a ton of clothes.” He merely described the blue jacket as part of her general walking ensemble, along with a variety of T-shirts, blue jeans, shorts and white tennis shoes she favored. Later, he told detectives he had looked in the laundry room where Sandy normally changed for her early-morning walks to avoid waking him, and because he did not see the jacket or other walking clothes in there, he surmised she had worn them that last night. In subsequent reports, the detectives simply wrote those were the clothes Sandy wore, despite Pat’s expressed uncertainty. But it is clear that, at the outset, Pat simply guessed at what Sandy had worn for her walk when she disappeared.