THE DNA WARS : Touted as an infallible method to identify criminals, DNA matching has mired courts in a vicious battle of expert witness.

THE AMBUSH WAS WAITING for Prof. Laurence Mueller long before he arrived at the courthouse with his bar charts and slides and other hieroglyphs of science. You could see it in the deputy D. A.’s bear-trap smile, in the pile of dog-eared files he had amassed on Mueller’s past courtroom performances, in the patiently highlighted inconsistencies ferreted out of those transcripts. The prosecutor and his experts were even making book--inaccurately, it turned out--that the good professor would wear his trademark tweed blazer, a coat that announces academe faster than any resume.

Mueller, a professor of evolutionary biology at UC Irvine, had been hired on behalf of an accused rapist to testify on the potentially eye-glazing subject of population genetics as it applies to that Wunderkind police technology, DNA “fingerprinting.” With such a topic, his main concern should have been keeping the jury awake.

He needn’t have worried. By the time Mueller left that courtroom in Santa Ana last spring, science, the presumptive subject of the day, had taken a back seat to more personal, and far more heated, topics. In short order, Mueller was branded a hired gun devoid of principles because he has earned more as an expert witness in DNA cases than as a university professor. As for his opinions, Mueller wasn’t just proclaimed mistaken, he was accused of attempting to “lie with statistics” and of making deliberately erroneous conclusions with false data, a cardinal sin in the scientific world. (Mueller and his colleagues say his only sin was disagreeing with the government’s experts.) Finally, Mueller was mocked for daring to apply lessons from his experiments with fruit flies to human genetics, even though the lowly fruit fly, with its prodigious birth rate, has been a staple of genetic science for most of this century.

The two-fisted, in-your-face cross-examination had jurors on the edges of their chairs and the defense lawyer itching to fling a similar attack at the prosecution’s three experts. Which he did, with the same sort of biting accusations Mueller had withstood, charging the experts with ignoring data to reach false conclusions.

Funny thing was, Mueller declared the experience one of his milder courtroom bouts to date. More often than not, the verbal body blows get a lot nastier. “It was heated, yes, but not as bad as some,” Mueller said. Later he added, “And I’ve taken a few other jackets out of mothballs.”

GENETIC PRINTING, A ONCE seemingly unassailable technology for identifying criminals through their DNA, has been stripped of its luster by a series of furious academic debates, vicious legal battles and highly personal attacks on scientists nationwide. Mueller’s experience has become the rule, not the exception, and people on either side of the equation say that both justice and science are the ultimate casualties of this war.

DNA printing, it turns out, though an undeniably powerful tool for the police, is not the nearly infallible magic bullet for solving crimes its proponents once claimed. Bottom line: Much of the time this method of comparing a suspect’s DNA to blood, semen or hair found at crime scenes works, sometimes it doesn’t. And the close calls--when the spotty, ink-blot-like patterns of deoxyribonucleic acid are blurred or faint--become so subjective that their utility is questionable. Matching a suspect’s DNA to evidence from a crime scene becomes an art, not a science, with lives and liberty hanging in the balance.

This simple realization has created chaos in the courts. Judges and juries with no scientific training are being asked to decide which scientists are right: those who believe DNA printing works well, or those who believe it is fatally flawed. Each side is sincere. Each cites reams of data. Each boasts impeccable resumes, and equally impeccable reasons for slamming those who disagree. So how do jurors know whether a complex technology is being used to convict the guilty, or to railroad the innocent? They can listen to hours of testimony and still not know. Innocents might be sent to Death Row, while killers are set free. It all depends on which side has the cleverest lawyers and the most persuasive scientists. Across the country, the confusion is so great that one judge can allow DNA printing as evidence while another, in the same state or even the same courthouse, could rule that the technology is still too unreliable.

And so, perhaps inevitably, a strange thing has happened as lawyers jockey for the sympathies of juries and jurists. Instead of the science, the scientists themselves are being challenged with unprecedented venom. Reputations have been tarnished, accusations of government conspiracies and defense lawyers’ cabals have been thrown onto the fire and bitter rifts among researchers have arisen, making it increasingly difficult for them to work together to resolve the very debate causing the courtroom combat.

The reason for this lies in a longstanding legal principle that judges may admit new scientific evidence in the courtroom only when there is no significant debate within the scientific community over its validity. That means government prosecutors and experts who want to use DNA to convict people cannot simply disagree with scientists who question the reliability of genetic printing. They must discredit and denounce the critics, questioning their motives, ethics and abilities to show that there really is no legitimate debate. Defense lawyers have responded in kind, attacking with equal zest.

Scientists accustomed to the sterility of the laboratory and the decorum of university conferences have been left blinking in anger, and sometimes fear, at the withering attacks. Allegations of perjury and hidden conflicts of interest, smirking remarks about sexual orientation, insinuations about tax audits and immigration status--nothing is off-limits. Many researchers now simply refuse to testify rather than face being shredded on the stand.

“It’s not about science,” says a rueful Mueller. “They are interested in dirty laundry.”

“I’m not ashamed,” counters Alameda County Deputy Dist. Atty. Rockne Harmon, a nationally prominent expert at using DNA evidence in court who has been equally prominent in efforts to discredit Mueller and other scientists critical of the technology. During charitable moments, he calls Mueller a “knucklehead.” “A dangerous criminal,” he says, “could be set free to rape, rob or kill people.”

In short, the battle over DNA fingerprinting has become the most entertaining and bewildering legal spectacle around. With jurors and judges unsure just whom to believe, resolution is nowhere in sight.

BY THE END OF LAST YEAR, THE FBI HAD PERFORMED 4,000 DNA COMparisons in criminal cases, with private, state and county labs adding thousands more. In about a third of those cases, the tests cleared the suspect. But hundreds of people were prosecuted using DNA evidence in more than 40 states.

Most of these defendants were convicted, a testament to the power of this new scientific evidence. But even when DNA tests were ruled inadmissible, the suspects usually were still found guilty. In part, that is because prosecutors chose slam-dunk cases for their first DNA trials, under the theory that obviously guilty individuals were less likely to pursue appeals of the DNA portion of their cases. Only now are numbers of cases beginning to appear in which DNA is the key piece of evidence, the only link between a suspect and a crime.

In the earlier mold, one case stands out. In New York, a handyman by the name of Joseph Castro was accused of fatally stabbing a pregnant South Bronx woman and her 2-year-old daughter in February, 1987. Lifecodes Corp. of Valhalla, N.Y., one of three private labs in the United States that does forensic genetics testing, was hired by prosecutors to do a DNA analysis. Lifecodes’ tests said the DNA from a tiny bloodstain on Castro’s wristwatch matched the woman’s genes, and that the odds against the blood belonging to someone else were 1 in 100 million.

But the Castro case represented the first full-blown challenge to DNA evidence. Castro’s lawyers amassed a team of prestigious scientists, who examined Lifecodes’ test results and found numerous errors. Even one of the prosecution experts, after conferring with the defense scientists, reversed himself and said that the vaunted 1 in 100 million match was wrong. The judge hearing Castro’s case kicked the DNA evidence out.

Yet prosecutors also proclaimed a victory: Castro later pleaded guilty, admitting that the blood on his watch really had come from the victim. Prosecutors say the Castro case, in which the tests were accurate but ruled inadmissible, raises the specter of criminals’ going free. Had there not been other, damning evidence in the case, they say, a double murderer would have walked. (In fact, although no national statistics are kept, experts on both sides of the DNA issue could recall only three or four defendants who won acquittals after DNA evidence was excluded.) Defense attorneys counter that bogus DNA evidence could just as easily convict an innocent person.

The prosecution of David Hicks, meanwhile, demonstrates another extreme--a case where DNA was the pivotal evidence despite problems with the test results. A young, unskilled laborer now on Death Row for the murder-rape of his grandmother in rural Freestone County, Tex., Hicks lacked the expert witnesses and skilled attorneys Castro had. The blurry DNA prints obtained in his case, and the 96 million to 1 probability of guilt announced by prosecution experts, again from Lifecodes, were barely challenged. The one expert hired for Hicks’ defense, former UC Irvine geneticist Simon Ford, said he got such limited and late access to lab data that he was unprepared when he testified. Yet the defense attorney appointed to represent Hicks declined to ask for more time.

After Hicks was convicted in early 1989 and sentenced to death, Ford and other scientists, Laurence Mueller among them, reviewed the Lifecodes lab work and concluded that the DNA match was improperly done and vastly overstated. An accurate analysis, Ford estimates, would show not a 96 million to 1 probability that Hicks was the killer, but something on the order of hundreds to 1. Furthermore, in Furney-Richardson, Tex., population 300, where Hicks was born and lived, many residents are genetically related. Families have lived there, married their neighbors and passed around the same genetic patterns, the same DNA, for generations. Therefore, people in the community are far more likely to share DNA prints with Hicks--and the killer. Indeed, relatives of Hicks were suspects early in the case, raising an even greater possibility that the killer’s DNA pattern could appear to match Hicks’, yet not be his.

Hicks’ jury never heard any of this because Ford never had a chance to do his analysis before the trial. Had they known, Hicks might easily have been acquitted, since there was little more than circumstantial evidence against him and no eyewitnesses, says William C. Thompson, the attorney and UC Irvine professor now helping Hicks to find legal representation for his appeal. Because of the certainty the 96 million to 1 finding implied, “the jurors were told that there was no way anyone other than Hicks could have been the killer,” Thompson says. “That simply is not true.”

Hicks’ prosecutor, Robert W. Gage, said in a letter to The Sciences magazine that he was convinced Hicks received a fair trial. Gage also asserted that Ford and Thompson, who together have written articles on DNA printing, have less than altruistic motives behind their criticism. They “profit by traveling around the country and testifying for desperate defendants about the pitfalls of DNA identification,” Gage claimed, noting that Ford and Thompson, thanks to their DNA work for defendants, are known to prosecutors nationwide as the “Combine from Irvine.”

Both Ford and Thompson say such remarks are examples of widespread harassment against DNA critics. Thompson, who is married and lives with his wife and children near the UC Irvine campus, says opponents are spreading false rumors among the DNA litigation set that he and Ford are lovers using their DNA legal fees to build a home in Laguna Beach. He has also endured bar complaints filed by prosecutors in New Mexico and California, accusing him of conflict of interest for publishing academic articles on cases in which he had been a lawyer. The complaints were ruled groundless, he says.

Ford also complains of prosecutorial harassment and, though still active as a consultant in DNA cases, now avoids testifying in court. A British national, Ford was on the witness stand in an Arizona murder case two years ago when an FBI lawyer abruptly asked about his visa status. Ford was then trying to obtain a permanent resident “green card.” Though the judge quickly silenced the attorney for straying into irrelevant areas, Ford saw it as a veiled threat from a federal official.

“It definitely was chilling,” Ford recalls. The judge in the Arizona case eventually dealt a stinging blow to the FBI, by finding that its DNA methods were not generally accepted by the scientific community and so could not be used in court. A state appeals court is currently considering whether to ban DNA evidence throughout Arizona.

In Texas, however, David Hicks remains on Death Row, convicted on similar evidence, insisting he is innocent, but garnering few sympathetic ears. Thompson believes Hicks might well be an innocent man convicted wrongly by his own DNA. But given the current conservative climate in appellate courts, Thompson concedes, stopping Hicks’ date with lethal injection will be an uphill battle. Yet the question his case poses is a haunting one: How can evidence be too unreliable in Arizona, yet be used to put a man to death in Texas?

TO KEEP DEFENSE EXPERTS AT BAY, AN INFORMAL NETWORK OF PROSEcutors and scientists who wholeheartedly support DNA printing has sprung up across the country. Rockne Harmon, the Alameda County deputy district attorney, is its unofficial clearinghouse. To help discredit defense experts, network members regularly fax articles, transcripts and other tidbits to one another, including, at times, unpublished papers that DNA experts have tried to keep confidential.

Researchers critical of the technology say they have been pressured to alter or withdraw scientific papers from publication. When pressure on individual scientists has failed, government officials and the experts they employ have lobbied publications, including the prestigious journal Science, to reject articles that challenge DNA fingerprinting. In one highly publicized case last December, Science--after pressure from DNA proponents--published an unprecedented, simultaneous rebuttal to an article criticizing the government’s theories on DNA printing.

“I never expected that the government would attempt to interfere in my scholarly activities or publications,” wrote Daniel L. Hartl, a co-author of the Science article and a professor of genetics at Washington University School of Medicine in St. Louis, in a sworn affidavit. James Wooley, a federal prosecutor from an organized crime strike force in Ohio, obtained the article before it was published, then telephoned Hartl, who had been a defense witness in a case Wooley prosecuted. According to the professor, Wooley “proceeded to badger me for almost an hour asserting that the article would do incalculable harm to government prosecutions and the criminal justice system. . . . I was particularly disturbed when Mr. Wooley . . . asked me whether I was afraid of having my taxes audited.”

Wooley denies any attempt to stifle Hartl’s, or anyone else’s, academic freedom, saying he may have made some sarcastic remarks toward the end of his telephone conversation with Hartl, but that the scientist was being paranoid.

Perhaps more than anyone, Laurence Mueller has been a favorite whipping boy in the DNA debate. Deputy Dist. Atty. Harmon has written harsh letters questioning the quality and accuracy of his testimony, not only to various scientific journals to which Mueller has submitted articles, but to his department chairman and the UC Irvine chancellor. Editors at Science even informed Harmon by mail that they had rejected an article by Mueller before they told the scientist of their decision. “It sounds like they respect me more than they do him,” Harmon says.

The government is not the only side getting down and dirty: Defense lawyers active in DNA cases also have their own network of litigators and scientists, and the government’s “pro-DNA” experts have begun to complain that they, too, are being harassed by harsh attacks on their integrity.

In U. S. vs. Yee, an Ohio murder case, both sides of the DNA war pulled out all stops. Last spring, New York defense lawyers Barry Scheck and Peter Neufeld filed a motion for a new trial, charging the government with conducting a national campaign to harass and stifle DNA critics. At the same time, the two attorneys launched an attack of their own: The motion accuses two prominent Texas genetics researchers of warping their scientific opinions in favor of DNA printing in order to garner $500,000 in Justice Department grants.

Scheck and Neufeld’s broadside also accuses the FBI’s chief expert in DNA printing, Bruce Budowle, of citing a nonexistent study to support convictions in the Yee case. The lawyers have demanded an investigation of this alleged perjury by the FBI, the nation’s preeminent purveyor of DNA printing. Budowle did not return repeated phone messages, but his boss, FBI Assistant Director John W. Hicks, defended the DNA expert, saying that Budowle knows that teams of defense lawyers and experts will scrutinize every word he says on the witness stand. “He knows not to say something dumb.”

Not even Hicks is immune from attack: Yee lawyers say he tried to destroy evidence that could aid defendants in DNA cases--allegations that Hicks hotly denies, although government memos do show that he wanted to destroy certain files that might have contained information on FBI lab errors. They also criticized Hicks for attempting to influence a recent report by a committee of the prestigious National Academy of Sciences, which in April issued a qualified endorsement of the use of DNA fingerprinting in court, but rejected certain crucial techniques used by the FBI. Hicks calls the allegations raised in the Yee case “a witch hunt.”

FINDING COMMON GROUND BETWEEN THE OPPOSING SIDES HAS been nearly impossible, in part because of the way DNA evidence was first presented to the courts and the public: as a magic bullet.

The technique first was used in 1985 in Great Britain, when blood was taken from more than 4,000 men to identify a rapist-murderer who had killed two 15-year-old girls in two neighboring villages. The DNA fragments extracted from these blood samples were then compared to DNA from semen found in the victims; police found their man when he tried to get someone else to contribute a blood sample for him.

Proponents of the technique claimed that the genetic comparison was so exact that the odds of a false match were 1 in hundreds of millions. In other words, they said, the odds of a mistake were so minuscule that DNA was as good as a fingerprint.

By 1987, the technology had crossed the Atlantic, with three private laboratories offering DNA analyses in criminal cases. The FBI soon followed with labs in Washington and Quantico, Va., and various state and local government DNA laboratories came on line in the next two years. DNA printing became a growth industry.

Its appeal was enormous, especially in rape and murder cases. Previously, forensic scientists charged with finding evidence at crime scenes could only compare blood types, or a somewhat more refined analysis of blood enzymes. Semen or blood found in the victims or at the crime scenes was collected and compared to blood from suspects. If there was a different blood type, a suspect was exonerated. If there was a match, however, it often proved little; millions of people share the same blood type or enzyme.

“Before DNA, if we got a figure of 1 in 4, maybe 1 in 10, we were lucky. One in 100, we were ecstatic,” says John Hartmann of the Orange County sheriff-coroner’s DNA lab, one of the most respected in the nation. “It was not very discriminating.”

Then, suddenly, a quantum advance occurred. Compare DNA, proponents of the new technology announced, and you could be sure you had the right culprit. To men like Hartmann, who genuinely agonized at the thought of criminals’ going free because their lab work was too inexact, DNA printing represented a law-enforcement “home run.” Now rapists who might have gone free because of the vagaries of eyewitness testimony or the lack of hard proof could be prosecuted. Serial murderers who left no living witnesses would no longer prey on society with impunity. Their DNA, with the help of the knowing, dispassionate men in white lab coats who analyzed it, would answer the questions of guilt and innocence. And the odds that they might be wrong would be 1 in millions, maybe billions. It almost seemed too good to be true.

It was.

Even the term genetic “fingerprint” turned out to be a misnomer, granting an unwarranted aura of infallibility to the technique. Now many jurisdictions prohibit the term’s use in court out of concern that it misleads juries, using instead DNA “profiling” or “typing.”

Jurors can look at a projection of conventional fingerprint comparisons and say, yeah, they really do look alike. Furthermore, it is uncontrovertible that no two people possess the same fingerprints, not even twins.

Neither is true with DNA prints.

Unlike fingerprints, a person’s entire DNA is not examined in the technique--current technology is too primitive. Instead, scientists analyze only a few areas of the long, complex chains of deoxyribonucleic acid that contain the genetic blueprint for all organisms, bacteria to Homo sapiens. The regions of the DNA molecule that are studied, fragments called VNTRs, serve no known purpose other than linking other portions of DNA, like empty boxcars in the middle of a freight train.

When a rapist or a killer leaves semen, blood or hair at a crime scene, the DNA can be chemically removed and broken up into fragments in a solution of organic solvents and enzymes. The purified DNA that results is poured into an electrified gel, which spreads out the fragments according to their length. Afterseveral more steps, another chemical solution called a “probe” is added, which mates with a specific group of VNTRs, highlighting it with a radioactive tag. The final step uses radiation-sensitive film to record the VNTR pattern. The resulting “autoradiograph” bears a passing resemblance to a very sloppy bar code like that on most grocery labels. The crime-scene autorad can then be compared to a suspect’s autorad. Theoretically, different patterns mean innocence; a match means guilt.

Sounds a lot like fingerprinting. And proponents and critics agree that the technique is theoretically sound. But in practice, the autorads are occasionally hazy and incomplete, due in part to the decomposition that typically occurs before forensic evidence is gathered. The blurry images that can result may lead one expert to see a match where another sees none. Many matches are clear-cut; others are so subjective that the certainty DNA printing is supposed to convey simply vanishes. Worse, jurors and judges at times don’t know what they’re really looking at when the scientists start passing out the autorads. So one focus of the debate is whether a match can be declared reliably in each and every case.

The hottest battle, though, revolves around what it means when DNA prints do clearly match. Does a match really mean that only one person matches the criminal’s DNA print, or are there other people walking around with the same patterns? Proponents of the technology say the odds of more than one person bearing the same VNTR patterns are extremely small. Critics say prosecutors and their witnesses are vastly overstating this certainty level.

To understand this aspect of the debate, you have to understand how the chemical “probes” work. There are several kinds, and each probe examines a different grouping of genetic “boxcars”--the VNTRs. According to FBI population studies of several thousand people who gave blood samples, one VNTR grouping may occur in 1 out of 100 people. Another grouping may occur in 1 out of 17 individuals, another, 1 in 2,000. By using three, four or more such probes, then multiplying the odds together, incredibly small probabilities are generated, anywhere from 1 in 100,000 to 1 in billions. Thus, the image of DNA printing as magic bullet was born.

In the Hicks case in Texas, for example, Lifecodes declared a six-probe match in calculating the 96 million to 1 probability that he raped and murdered his grandmother. In theory, this should be irrefutable evidence. But defense witness Ford’s analysis showed that four of the six VNTR patterns contained in a vaginal swab of the victim could have come from the grandmother’s cells, not the killer’s--there was no way of telling them apart in this case. As relatives, Hicks and his grandmother had identical VNTR groupings in four places. Using only the two VNTR groupings that must have come from the killer and that do match Hicks, the odds of Hicks’ being the guilty party, Ford says, are a few hundred to 1, nothing close to the level of certainty represented to the jury that convicted Hicks. By Ford’s estimate, millions of others could have committed the crime, including many people in Hicks’ insular hometown.

The possibility that such communities might contain people with common VNTR patterns is another hot point of contention in the DNA wars. Defense attorneys recently sought to embarrass one of the government’s most prominent DNA experts, Kenneth Kidd, a professor of genetics, psychiatry and biology at Yale University, for claiming that a four-probe match between two different people was virtually impossible. This claim was refuted--with Kidd’s own research on an isolated Amazonian Indian tribe called the Karitiana, in which about a third of the 54 people tested had identical DNA patterns for four different probes. (Hicks was sent to Death Row on the basis of two distinct probes.) Once confronted with his own data, Kidd and prosecutors dismissed the inbred Karitiana community as an aberration that does not apply to the much larger gene pool of U.S. populations.

“That’s what they say now,” Thompson argues. “But before, they said it couldn’t happen at all, anywhere.”

The DNA critics say such findings dictate that more conservative probabilities should be used to explain to juries what a DNA match means. Prosecutors resist this, saying the FBI and other labs already build in error factors that give defendants the benefit of the doubt. Obviously, they also oppose change because less dramatic odds leave too much room for doubt about a suspect’s guilt. The magic bullet turns into a blank.

The recent National Academy of Sciences report, however, urges somewhat more conservative numbers than the FBI and other labs generate--striking a rare compromise between the two DNA factions. It remains to be seen if either side adopts the report’s recommendations. So far, the FBI is resisting change, while defense lawyers lobby for probabilities more in their favor. A series of recent appellate court decisions against DNA evidence, including one in Massachusetts and two in California, cite the academy report and side with the critics.

“Really, when you get down to it, there’s not much difference between a million to 1 and 100 million to 1, or a million to 1 and a 100,000 to 1, which is really what they’re arguing about,” Orange County’s John Hartmann says. “Either way, we’ve hit a home run.”

IF SUCH ARGUMENTS WERE all there was to it, the viciousness that marks DNA court battles probably wouldn’t have erupted. Each side merely would amass its experts, let them testify, and the judge and jury would decide. But the battle also revolves around a legal precedent set by a federal appeals court in 1923, the Frye Rule, which governs the admissibility of new scientific evidence in many, though not all, jurisdictions. The legal rule was first established to examine (and ultimately bar from court) lie-detector tests, but 70 years later, it provides an uneasy fit with the state-of-the-art technology of DNA printing. In DNA cases, Frye’s peculiarities make the intense attacks on scientists’ integrity all but inevitable.

According to the Frye decision, scientific discoveries should be admitted only when there is no substantial debate about them within the relevant scientific communities--in this case, geneticists, molecular biologists and statisticians. Under Frye, judges aren’t supposed to decide who is right in a scientific debate--they aren’t qualified--but merely whether a legitimate debate exists.

That explains the vicious attacks on the scientists, and why so many complaints of harassment have been generated. If government officials can halt articles and paint DNA critics as unethical charlatans, judges may decide there is no legitimate debate.

Naturally, defense lawyers respond in kind, attempting to prove the government experts are the real charlatans whenever they testify that there is no debate about DNA printing. “It’s trench warfare, no doubt about that,” Hartmann says.

Until recently, prosecutors have won most of the skirmishes: DNA evidence has been admitted in most cases, with decisive results. Only one state’s Supreme Court has ruled DNA inadmissible--Massachusetts--while one other, Minnesota, severely limits the way it can be used. Numerous other states, including California, are reconsidering its admissibility. The U. S. Supreme Court has yet to specifically rule on the issue.

At the trial court level, the results are more contradictory. In the same courthouse, one judge may admit a DNA test while another finds a legitimate scientific dispute exists, rendering the DNA evidence inadmissible. Justice, when it comes to DNA evidence, is unequal, varying from state to state and judge to judge. The U. S. 2nd Circuit Court of Appeals (with jurisdiction over New York, Vermont and Connecticut), for instance, recently approved the use of DNA evidence without lengthy admissibility hearings; at the same time, a District of Columbia judge barred prosecutors from using DNA printing in court.

Another example: last year, one division of the California Court of Appeal upheld DNA evidence in a Ventura murder case, which could have set a precedent for all of California. But a month before that ruling, Los Angeles County Superior Court Judge C. Robert Simpson Jr. ruled in a different murder case that DNA evidence was too questionable to be admitted. “There is a profound, significant and honestly held disagreement among these men of science,” Simpson wrote. Because that case developed after the Ventura prosecution, Simpson’s decision was not undone by the appellate court’s embrace of DNA evidence. The two at-odds rulings stand, their illogic intact.

In August, a different division of the state Court of Appeal, citing the national academy report, ruled that DNA evidence was not admissible. Now the state Supreme Court has been asked to rule definitively.

An Orange County rape trial earlier this year, in which Mueller and Kidd were lambasted, shows the kind of balancing act juries must perform to deal with the attacks, disputes and contradictions. Frank Lee Soto was charged with raping his neighbor, an elderly woman in Westminster. The only hard evidence against him was a DNA test of his blood--he provided it voluntarily--that showed a clear match with the rapist’s semen. (The victim, too ill to testify, told police Soto didn’t do it.)

Soto’s attorney, Paul Stark, fought bitterly with Deputy Dist. Atty. Dennis Bauer over what the DNA match meant. Bauer, with Yale’s Kidd and several other of the most prominent pro-DNA printing experts in the country, said the odds were 189 million to 1 that Soto was the right man. Equally prominent experts suggested the numbers should be more favorable to Soto; in particular, biology professor William Shields of the State University of New York at Syracuse said the correct number was between 65,000 to 1 and 23,000 to 1. Given Orange County’s population, that meant 17 other potential suspects were running around, five times that in all of Southern California.

Both sides sounded certain and sincere. Both had the charts and the numbers and other paraphernalia to support their positions. Both sides impugned the abilities and motives of the opposing experts in typical vitriolic fashion. There was literally no way for the jurors (or the judge, for that matter) to know whom to believe.

Doing what the scientists refused to do, the jurors sought compromises. Their statements after the trial showed that they accepted the smaller defense numbers put forward by Shields, and so convicted Soto of the lesser charge of attempted rape. Had Soto not been a neighbor, they said, they would have acquitted him. But 65,000 to 1, coupled with the fact that he lived next door, was enough to cinch the case. Soto received a three-year prison sentence, but, because of the uncertainties revolving around the DNA issue, he is free on bail while he appeals.

Hartmann, whose Orange County lab performed the tests on Soto’s blood, applauded the jury’s decision that DNA evidence alone is insufficient to convict. His attitude is among the most reasonable of any expert in the field and though he is ardent in his support for DNA printing, he does not believe that it should on its own determine anyone’s fate: “If I was on that jury, I would have had a real problem with it, too. . . . No one should be convicted solely on the basis of DNA evidence, not even a billion to 1. If there was no other evidence, I would vote not guilty, too.”

BUT FOR MOST OF THE COMBATants, the DNA debate has assumed the aspect of a religious argument, each side certain it is right, each incapable of seeing the other’s point of view. Voices of compromise are drowned out.

There is one unique case in Virginia that calls into question the sincerity of these head-butting positions. Joseph R. O’Dell III was convicted of a vicious and sensational murder in Virginia Beach in 1985. O’Dell was seen at the bar where the murder victim had been on the night of her death, though no one saw them arrive or leave together. Later, police seized bloody men’s clothes in O’Dell’s garage after they received an angry call from his girlfriend. O’Dell, who had a robbery conviction and other legal run-ins behind him, told police his shirt and jacket got bloody from a bar fight that night, but detectives figured they had found their man. Lab results seemed to confirm their opinion. The case predates DNA printing, but less sophisticated blood enzyme tests showed that the blood on O’Dell’s clothes was consistent with the victim’s blood. O’Dell was convicted and sentenced to death.

Since then, with the advent of DNA testing, his appellate lawyers petitioned the court for new tests. The result: the more refined DNA technology shows the blood on O’Dell’s clothing cannot be matched to the victim’s, according to briefs filed in his appeal. Key evidence against O’Dell apparently has been invalidated, a case where DNA printing seemingly exonerates rather than convicts.

But O’Dell remains on Death Row. He has been unable to win a new trial. Prosecutors, who in other courtrooms have championed the use of DNA evidence against defendants, are so convinced of O’Dell’s guilt that they have opposed its use in his case, even as defense attorneys want the test results admitted.

The prosecution is winning so far, keeping the Virginia appellate courts from considering the DNA evidence in O’Dell’s case by focusing on a technicality. O’Dell’s out-of-state lawyers filed an appeal petition that lacked a table of contents and a memorandum of facts summarizing the case (a formulaic introduction to the appeal). The prosecution seized on this procedural gaffe as a means of throwing out the entire appeal. By the time the defense lawyers noticed the oversight, a filing deadline had passed by three days, and the Virginia Supreme Court refused to consider the case.

Recently, three justices of the U. S. Supreme Court suggested that the Virginia courts ought to think again. Although the full court declined to hear O’Dell’s direct appeal, Harry A. Blackmun, John Paul Stevens and Sandra Day O’Connor decided that putting a man to death because his lawyers forgot a table of contents was going too far, especially if he might be innocent. In an unusual letter handed down in the case, they instructed lower federal courts in Virginia to give careful consideration to O’Dell’s next appeal, a habeas corpus petition to be filed this year, “because of the gross injustice that would result if an innocent man were sentenced to death.” And there it stands.

O’Dell remains on Death Row, waiting for his case to be heard, even as prosecutors across the country continue to use DNA evidence to send him companions.