If the Genes Fit,
How Do You Acquit? O.J. and Science
Andrew Ross
I was only in Los Angeles once during the whole O.J. circus, but the memory of a brief encounter in the cocktail lounge of a Santa Monica hotel in August 1994 has stayed with me. A bar bore was getting cranked up, and soon he was broadcasting loud and clear on the topic du jour. He was a white businessman, whom the bartender seemed to know all too well and wished she didn’t. He was claiming a special knowledge about football players on the basis of having once served on the board of trustees of an East Coast university. Without much egging on from his audience of two, one bluster led to another until he unveiled his theory that, of all football players, the success of running backs on the field depended less on brainpower than on their killer instinct, and that this disposition was probably genetic. In fact, he was sure this was the key to the O.J. Simpson case. When pressed to elaborate, he ventured that the scientists could prove this kind of genetic theory in the courts, but they would not be allowed to do so in the upcoming trial. At this point, because I had to leave to meet a friend, I discreetly expressed my sympathies to the bartender who would now bear the brunt of these pronouncements alone. What I thought I had encountered in the bar was a new kind of concoction—of half-truths about the links between genetics, law, and crime—that may be more potent and widely consumed than anyone would like to believe.
Back home in New York two weeks later, another friend—an O.J. trial junkie—who shares some of my interest in genetics sent me a clipping of a New York Post headline that read “Murder in O.J.’s Blood.” The clipping confirmed my impressions in that throat-grabbing way tabloid headlines have when they tell us “this is where you live, here and now.” The headline, no surprise, bore only an oblique relation to the newspaper story, which soberly recounted that preliminary results of two different DNA tests had matched O.J.’s blood with blood found leading from the crime scene. But the headline’s telegraphic announcement suggested something else, much closer to my bar bore’s woozy story about the genetic basis of criminality.
What was happening in August 1994 to explain this kind of nonsense? In “the Simpson matter” (as Judge Lance Ito would famously come to refer to the trial, day after day), the prosecution and the defense teams were both being assembled, shaped by the case that each was going to present. It looked as though the prosecution’s case would rely heavily on DNA profiles of blood left at the crime scene and on various objects—the bloody glove, the Bronco—to identify Simpson as the murderer, and, to a much lesser extent, on the demonstration of motive provided in his history of domestic violence. It had just been announced that the defense would bring on board Barry Scheck and Peter Neufeld, directors of the Innocence Project, well known for using DNA testing to exonerate over a dozen clients who had previously been convicted and imprisoned. These attorneys would play a crucial, and perhaps decisive, role in casting reasonable doubt on the DNA evidence in the course of the trial. The prosecution, in turn, would call on the expert testimony of Rockne Harmon, George “Woody” Clarke, and other strong advocates of DNA forensic technology. In August, then, both sides were preparing for what the press was referring to as “the DNA Wars,” scheduled to begin, and potentially end, in the pretrial hearings that would determine the admissibility of DNA evidence. There was reason to believe that some, if not all, of the forensic techniques would be considered inadmissible in the California courts.1 This is what I think my bar bore might have had in mind when he speculated that the scientists would not be allowed to prove his theory in court.
To get to the pulpy core of his theory, however, we need to step outside of the Simpson matter and look briefly at the profound impact that modern genetics has had on public consciousness in the last fifteen years or so, to the degree that it has become routine fare in barroom conversations. In that period of time, it has become common to refer to genetic causation for virtually every aspect of human behavior—from medical disorders to ethnic traits and personal conduct. Science reporting has encouraged the view that such genetic links are simple and direct, and in popular culture it is currently acceptable to say there is a gene for almost everything, including the urge to make generalizations of this sort. Much of this attribution is droll, and therefore it mocks even as it reflects the exalted status of the Holy Gene.
Even so, the popular allure of biological explanations for social behavior is encouraged not only by biotech companies but also by scientists and policymakers who have helped, directly or otherwise, to shape the attack on collective social responsibility in general and the welfare state in particular. Absolving society of all blame or responsibility for problems that can be seen as genetically determined has been a significant element of the conservative crusade to redefine social problems as a matter of individual predisposition.2 In this view, gene therapy is more cost-efficient than social programs.
Of all the qualities that have been “linked” to specific genes—selfishness, shyness, adventurousness—the association with criminality is the most notorious. Talk about “natural-born killers” with “bad seed” or “tainted blood” is almost universal folklore, but its scientific basis has been consistently explored for over a century, from Dr. Professor Cesare Lombroso’s attempted criminal typology on the basis of physical characteristics in the late nineteenth century to theories of the criminal chromosome (the extra Y chromosome in males that accounts for their aggressivity) that surfaced in the 1970s. Ever since then, ideas about the genetic basis of criminality have been riding the long wave of resurgent biologism into the institutions of social and legal policy-making. The currency of these ideas has been tied to the fortunes of the eugenics movement over that same period of time, rising with Galton in the late nineteenth century, collapsing in the aftermath of Nazism, and surfacing again in new forms primarily in the U.K. and the U.S. through the developing paradigms of sociobiology and evolutionary psychology in the last two decades of the twentieth century.
The concept of genetic criminality has contributed specifically to the social psychology of racialization. Most studies of genetic criminality survey incarcerated populations, and since African Americans are incarcerated at a rate ten times that of Caucasians, it is hardly surprising that U.S. studies draw bogus conclusions about links between race and criminality.3 Given the high public tolerance of the U.S.’s racially skewed incarceration rates, it is easy to imagine the appeal to racists of such conclusions, casually linked, in turn, to other spurious ideas about genetic causality and race, such as the link between race and intelligence.
By the end of the summer of 1994, there were very few public scandals that could have dislodged O.J. from the forefront of North American public conversation. In some circles, the publication of The Bell Curve, by Charles Murray and Richard Herrnstein, came very close. The cant about genetic links between race and IQ dredged up by Murray and Herrnstein was designed to feed intravenously the antiwelfare policies of the moment. Their ideas about the decline of the gene pool and the racialization of intelligence in the new cognitive hierarchies of information society seemed to be borne along by some autonomous principle of social motion, at a time when popular opinion is increasingly receptive to arguments that appeal to genetic authority.
The public debate about The Bell Curve lasted for several months, and, while it did not make much news in the tabloid press, it occupied many of the same media pages as the O.J. trial. However, my extensive data-base search for stories that mentioned both O.J. and The Bell Curve during the relevant period and shortly thereafter came up with very few matches and none at all that touch on genetics. This discovery surprised me. Both The Bell Curve and the DNA typing in the O.J. trial employed a controversial process of testing, appealed to racial or ethnic categories as classified by genetic scientists, and rested to some degree on assumptions about how genetics can be used to elucidate legal or social quandaries. Why then was discussion of one field of genetics so segregated from the other in the public media? Surely there was some reason other than that The Bell Curve appealed to a behavioral, and the O.J. trial to a forensic, field of genetics.
Was it because one had been marked as “contaminated” by its racist associations, while the other was seen as clean, at least where race is concerned? Given that DNA forensics have evolved in response to the needs of a law enforcement system with a systematically racist record, only those immune to skepticism would view this as a credible explanation. Indeed, it was partly on the basis of that racist record that members of the O.J. defense team argued their case for the incompetence of the LAPD criminologists and the racial biases of one or more of its detectives.
Or was it because The Bell Curve was perceived as a soft application of molecular biology, and DNA typing as harder, and more incontrovertible? Perhaps, but then one of the big media stories about the DNA wars was that the forensic methods were highly contestable. Ever since the 1989 case of People v. Castro,4which successfully challenged DNA typing as an infallible source of legal evidence, thiscomplex technology has been fiercely disputed in courts and in scientific and law journals, and it was the object of bitter contention in the O.J. trial.
The reasons for the separation of interest in O.J. and The Bell Curve do not lie within molecular biology itself. Instead, they arise from the needs of those institutions—in education, law, medicine, commerce, the military, and government—that serve as clients for customized kinds of scientific knowledge. Specialized scientific fields are habitually made to order by powerful institutions or wealthy contractors or because they respond to prevailing ideological needs. In almost every field, the division of labor within the scientific community is shaped by the managerial needs of social and economic elites. Subfields of scientific knowledge are social arrangements of labor (many in the late twentieth century are spin-offs from the military-industrial system). The division between different fields of genetics corresponds in part to the select interests of clients. Invariably, demand from one field shapes developments in another; DNA forensics, for example, relies upon population genetics for its identification of suspects. Division of fields, moreover, does not preclude their independent reinforcement of the powerful philosophy of genetic determinism.
When it is not directly converted into capital, specialized scientific knowledge is habitually sought out to intensify respect for institutions and their policies. Consider some of the respective outcomes of the O.J. trial and The Bell Curve. While very few scholars leapt forward to defend the scientific premises of The Bell Curve, few would doubt that its poisonous theses about alleged links between race and intelligence have had considerable impact on framing the public reception of major changes in social legislation, especially those relating to affirmative action policies. Indeed, it could be said that conservative policymakers are proceeding toward the conclusions favored by the book without relying overtly on the junk science offered by Murray and Herrnstein. By contrast, mainstream media wisdom held that the DNA evidence against O.J., while scientifically sound, was somehow disregarded by the jurors. This perception helped to fuel widespread disrespect for the jury system in the months after the trial, and may even have contributed to the zeal with which powerful politicians from the White House down took up the old game of judge-bashing. In the case of The Bell Curve, the science, even though it was discredited, still helped in some way to “explain” the policy-making directed against social programs. In the Simpson matter, more accredited science was called for to better calibrate the institutional fit between the forensics field and its legal interpretation.
Accordingly, the science establishment played its role in helping to restore post-O.J. public confidence in DNA forensics. In July 1996, the NRC (National Research Council, the operating arm of the National Academy of Sciences) published an updated report on DNA evidence that issued a clean bill of health for the forensic process: “The technology for DNA profiling and the methods for estimating frequencies and related statistics have progressed to the point where the reliability and validity of properly collected and analyzed DNA data should not be in doubt.”5 Among its recommendations, the report urged the professionalization and accreditation of all crime labs in accord with upgraded standards of performance and accountability. This resolve is consistent with the history of the development of criminological technologies, from fingerprinting to polygraph testing, voice printing, blood grouping, enzyme typing, and DNA profiling. Each has required its own credentialed field of experts, with a professionalized jargon, a supporting record of approval in peer-reviewed journals, and a bureaucratic system of accreditation with institutional authorities. When doubts are cast on the legal interpretation of such technologies, the response of the scientific community has been to increase funding in pursuit of more accurate data and to beef up the supporting legitimating network of professionalization.
From the time it was first challenged in U.S. courts in 1989, evidence for the fallibility of DNA profiling had been supported by indications that biotechnology companies had pushed hard for the legal admissibility of the tests, the FBI had intimidated scientists who were critical in print,6 and the “community of experts” responsible for validating the techniques involved many scientists who had themselves established a good deal of their reputation through lengthy and lucrative courtroom testimony about the value of the field to criminal jurisprudence. Aside from these more obvious conflicts of interest, the debate about the increasing use of DNA sampling and profiling involves a host of ethical concerns about the threats to genetic privacy posed by the expansion of the state’s DNA identification banks, compiled from compulsory testing of prisoners, soldiers, and patients by the Departments of Defense and Justice and other state agencies; the normalization of genetically defined population differences; the further sanctification of biological explanations of the truth; and the applicability of quantitative reasoning to courtroom procedures designed to respect principles and values other than those acknowledged by science. Despite its public visibility, and the unusually generous resources available to both the prosecution and defense, the O.J. trial highlighted very few of these concerns.
The trial involved an enormous amount of scientific testimony (10,000 references to DNA in the 50,000-page transcript), inordinately more than, say, the 1953 Rosenberg case, another great “trial of the century,” in which the prosecutors had to argue that the lay defendants understood enough physics to pass on the secret of the atom bomb to the Soviet Union. Predictably, the press found much of this testimony regarding DNA “boring,” and an obstacle to the unfolding drama of the trial. Some commentators, like William F. Buckley, found the trial itself to be an obstacle to those goals of justice that were clearly aligned, in his view, with the evidencepresented by the science. Writing in the National Review, Buckley lamented that “justice is dying … from the creeping immobilizations brought on in the name of civil liberties…. The only obstacle to the establishment of the guilt of O.J. Simpson is legal. The whole of the epistemological apparatus of the modern world—psychology, science, logic, reason—establishes that he is guilty. Only the law stands in the way of the application, paradoxically, of justice.”7
Voicing a standard conservative view, Buckley manifests his impatience with the legal system’s overzealous protection of the civil rights and dignity of defendants. He sees the cart going before the horse: the law should be driven by science—not vice versa. In framing the debate over scientific evidence in this way, Buckley was rearticulating an age-old concern that the jury system poses a substantial threat to the rule of law (especially manifest in the principle of jury nullification). He was also anticipating the torrent of public impatience with the jurors’ verdict in the O.J. case. At issue is the Federal Rules’ entrustment in the judgment of jurors, over and above those of experts and professionals in science and the law. DNA evidence has presented the most acute challenge to that principle in recent years because it has stretched the capacity of experts to believe that lay people can understand advanced science. If complex scientific testimony increasingly becomes a customary presence in thecourts, what are the prospects for continuing respect for lay judgment?
The fierce public backlash against the O.J. jurors was immediate, palpable, and sustained long enough for it to have congealed into a media “fact.” The main tenor of the hostility stemmed from the belief that a predominantly black jury had exonerated a wealthy black celebrity with a history of spousal abuse in defiance of Marcia Clark’s “mountain of evidence,” and had done so in retaliation not just for the racist record of the LAPD but also for centuries of “white justice.” The reactionary version of this belief was downright ugly, but the more liberal interpretation of the verdict as a healthy exercise in jury nullification may have been no less insulting to many of the jurors. When such allegations were not being made directly, related charges were often insinuated by questioning the jurors’ lack of education in the face of the cold, hard facts of the DNA evidence. On many a talk show, we heard again and again the claim that since smart people can increasingly find ways of evading jury duty, juries are inevitably made up of the dumbest folks in the community. This claim has clearly racist overtones if one considers that the substantial increase in minority voter registration (along with the dismantling of all-white juries who, unlike black juries, are statistically inclined to exonerate defendants of their own race) has brought more and more citizens of color into the jury pool in recent years. For example, a March 31, 1996, segment of 60 Minutes suggested that Francine Florio-Bunten—a white juror removed from the trial after Judge Ito received a letter alleging she was working on a book with a literary agency—had been set up by the defense team. The segment implied that Florio-Bunten was the most educated of the jurors and was therefore capable of understanding the DNA evidence—and of swinging the jury against O.J. Similar insinuations and suspicions abound in the massive media archive. The famous Internet list, alt.fan.oj-simpson, featured a wave of repulsive postings after the verdict; typical sentiments included a posting that read “Hate to say it, but I think The Bell Curve hit the nail on the head,” or another which confessed, “I don’t know if I can look upon blacks as anything other than a subhuman mob with an average IQ of around 80.”8How did thisbacklash affect the jurors themselves? At least three—Armanda Cooley, Carrie Bess, and Marsha Rubin-Jackson, who published their accounts in Madam Foreman: A Rush to Judgment?—have described their persistent discomfort at being under suspicion themselves. Such was their impression that, in addition to their resentment at the cameras monitoring the hotel hallways, these jurors imagined that the smoke detectors and fire alarm units in their bedrooms were audio and video monitors (Rubin-Jackson took appropriate measures: “One day I got up in front of it buck naked … and shook my bootie.”).9 This paranoia was heightened by the intimidating scene outside the courtroom on the day the verdict was announced. Confronting the jurors as they arrived were hundreds of police officers on horseback and in riot gear, while helicopters seemed to fill the sky above. In downtown Los Angeles, this display of a massed LAPD force is by no means a neutral sight, especially for its citizens of color. With the crowd outside considerably swelled and agitated by the announcement of the verdict, these three jurors had some reason to feel they were not only besieged in the courtroom but also in some physical danger: “Every time I looked there were more and more people and I just kept saying, ‘Oh my God, are they going to riot, are they going to get us?’” (MF, 176).
In order to throw off the reporters and cameras, the jurors were ushered out of the courtroom through the route reserved for the transport of prisoners (“It was like running a gauntlet. I felt like I’d committed a crime and I was on my way to being locked up” (MF, 7). This route involved the use of the elevator ordinarily reserved for convicted felons, surrounded by walls of graffiti both poignant and defiant, scrawled by men and women en route to detention cells. Once outside, they were transported to a secret drop-off point in a “full-size black-and-white-bus with barred windows.” Ordered to keep their heads down, some of the jurors were fearful at the sight of the full extent of the massed police force: “Why do they need so many policemen? Do they really think we’re going to be assassinated?” (MF, 9). No wonder these jurors felt as if they were at the wrong end of the justice system, not only then but in the months to come.
These stories and impressions are important to recount because they are a record of how and why jurors feel that their involvement with the justice system can bring a share of its (and the media’s) punitive power down upon them. Something is clearly wrong if the fulfillment of their civic duties renders them vulnerable in this way. Public bigotry will only intensify if the two chief allegations of the O.J. juror backlash—that they practiced reverse racism and that they were undereducated—are allowed to subsist. For the record, again, let us note what jurors had to say about these allegations.
Having lived “underground” for the period of sequestration in circumstances more akin to detention (with five-hour conjugal visits and routine searches of their belongings), these jurors seemed genuinely surprised to learn that the verdict was being interpreted as a racial message, especially since Simpson was perceived as living “in a white world” of privilege quite remote from their own black one: “You could say we were shocked … or outraged, that people would even think of us sitting there making decisions based on race. And it proved to me that they felt we had no intelligence whatsoever” (MF, 82–83). The lengthy DNA testimony elicited a similar impression: “When I got out [sic], I kept hearing more and more people saying there was only one graduate on the jury and so forth, so that indicates to me that they felt everyone else was illiterate …. I realize that if you have a Ph.D. and you’re talking to someone who does not have training in the field of serology or forensic science, you might assume that they’re not going to understand some of the basics. Of course, you’re not going to understand the total details of that field, but you don’t have to. Unfortunately, there’s no way to let people know you got it. You can’t just raise your hand and say, ‘Dr. Cotton, I understand what you’re talking about. Move on.’” (MF, 114).
As it happens, jurors rated Dr. Robin Cotton, from Cellmark Diagnostics, the Germantown testing laboratory, lowest among the scientists judged on their ability to communicate technical arguments: “She talked down to us like we were illiterates….
She talked down and when you talk down to people you tend to lose them” (MF, 114). On the prosecution side, Renee Montgomery and Gary Sims, fromthe California Department of Justice, and Dr. Lackshamanan Sathyavagiswaran from the LAPD, rated highly, as did Henry Lee from the Connecticut State Forensic Science Laboratory, and Barry Scheck, on the defense side. Collin Yamauchi, from the LAPD, who had cast doubt on Lee’s handling of the evidence, was perceived as covering up for the egregious errors in gathering and handling of evidence by LAPD criminalist Dennis Fung and his assistant Andrea Mazzola.
In the final analysis, these jurors reported that they had in fact understood the testimony relating to the DNA matches. With the exception of one particularly degraded sample (from the rear gate at the Bundy condo) that had been collected weeks after the crime and found to contain high levels of EDTA, a blood preservative used in labs, the jury, in its deliberations, had not questioned any of the matches of the blood samples with Simpson’s DNA type. Instead, it was the evidence of the LAPD mishandling and cross-contamination of the blood evidence that was relevant to establishing reasonable doubt.
Of course, no one is obliged to take the jurors’ comments out of court at face value, but to ignore them is to add insult to their perceived injury. As it happens, these comments reveal a good deal about the balance between expert and lay opinion that has been affected by the legal admission of complex scientific evidence.
One public perception that emerged from the O.J. trial was that the science was only as good as the police lab that processed the blood evidence—a “cesspool of contamination” in the view of defense witnesses. Barry Scheck argued, in a 1994 article about the Daubert decision [see note 1], and unceasingly throughout the trial, that laboratory error is the primary flaw in the DNA-typing process, and that a lab’s error rate should be considered a matter of admissibility, and not weight, as judges have been inclined to rule.10 In the O.J. case, Scheck could hardly have hoped for a more vivid illustration of his contention that laboratory error rates are substantially higher than the DNA profiling estimates that labs draw from trace sample evidence. As if in response to the unspoken injunction of the O.J. verdict to put the house of DNA forensics in order, the 1996 NRC report issued strict recommendations about forensic protocols and lab proficiency testing. Convened to resolve the legacy of uncertainties bequeathed by the controversial report of the 1992 NRC commission, this second NRC committee was not charged with the task of assessing the impact of DNA sampling upon the criminal trial process, nor with its much wider role in the state’s databanking of genetic information from a wide variety of convenience samples: criminal records, law enforcement officers, soldiers, paternity-testing centers, blood banks, hospitals. With respect to the former, the NRC committee confined itself to recommending “behavioral” (i.e., scientific) “research on juror comprehension” of evidence from DNA profiles. As is customary, then, the scientific focus on value-neutral fact was isolated from the legal and social concerns; “that’s not our department.” But in neither department is it possible, or socially prudent, to isolate facts from values in this way.
In recent years, genetic information has become a tool for decision-making in a variety of legal fields—torts, criminal, trust, and estate law—and is increasingly used as a defense on grounds of genetic predisposition. Despite its promise of absolute precision and irrefutable truth—its “aura of infallibility,” as a Massachusetts Supreme Court decision put it—DNA evidence is commonly introduced in the form of quantified probability. For example, in the Simpson matter, the odds of some of the blood samples matching any African American or Caucasian other than O.J. were estimated as high as one in 170 million, and one in 6.8 billion Caucasians in the case of the genetic markers matching Nicole and Ron. These astronomical odds, however much they varied from estimate to estimate (another put the Nicole odds, in the case of a particular blood sample, at one in twenty-one billion) conveyed the message that you “can’t argue” with such numbers. How could O.J. possibly be innocent, given these odds? Such statistics carry the patina of irrefutable truth in a manner that tends to outweigh other kinds of evidence, like those supporting the motive in the murder charge by reference to the history of spousal abuse. Indeed, these odds were frequently cited in the courtroom of public opinion as overwhelming confirmation that science had proved O.J.’s guilt, and that the jury had disregarded science and, in Buckley’s words, the whole of modern epistemology. These are false and perilous assumptions.
In a classic 1971 article, “Trial by Mathematics” (longbefore DNA evidence became a controversial factor in legal adjudication), Lawrence Tribesummarized the problems raised by the practice of using statistical methods to resolve conflicting claims in lawsuits. Even if it were desirable for the legal system to defer toquantitative reasoning, Tribe asserts that statistical proofs “decrease the likelihood of accurate outcomes” in a trial.11 The impact of introducing statistical evidence to a jury’s prior probability assessment of a defendant’s guilt has a distorting effect, rendering an inference of guilt that is much greater than the evidence warrants. The hard, quantitative evidence will dwarfother “soft variables” like impressionistic evidence, according to Tribe, and will indubitably warp the jurors’ obligation to weigh all the evidence evenly. If the statistical assessment is introduced early, it is difficult for jurors not to focus on these overimpressive numbers, and hence the presumption of innocence is often thrownout before defendants have had their full say. So, too, the probability values attached to variables, such as estimating the risk of a frame-up, or an error in testimony, are something that only individual jurors can assign, and so, to give mathematical proofs full credence, each juror would have to be able to compute their own complex equations, involving hard and soft variables, to guarantee the accuracy of outcomes. In addition, the authoritative weight of statistics harm the chances of a peer community’s accepting a defendant’s acquittal (in O.J.’s case, this community would presumably be his white neighbors in Brentwood and not downtown African Americans). Finally, Tribe argues that the use of statistics threatens to alter the entire character of the trial process itself, imposing standards unconvincing to the “untutored contemporary intuition,” making the legal system appear “even more alien and inhuman than it already does,” and undermining its responsibility to protecting the defendant’s rights as a person.
Given its “aura of infallibility,” many believe that the subsequent introduction of DNA evidence in the courts has only exacerbated these problems and has further eroded defendants’ rights, especially when the defendant’s resources are too meager to muster counterevidence. O.J.’s privilege in this regard is truly exceptional, but it proved, nonetheless, contra Tribe, that a successful defense can technically be mounted against an overwhelming array of quantitative evidence. The larger flaw with Tribe’s argument, however, is its assumption that quantitative evidence and reasoning arise at the outset from value-free knowledge. It is easy to conclude then that methodologies of scientific reasoning and the adversarial procedures of the legal system are difficult to reconcile. The one aims at isolating absolute truths that are irrefutable in any time or place, the other expresses the relationship between the individual and the state as defined by civil principles and rights that pledge respect for defendants as persons. As Marjorie Maguire Schultz puts it, “science deals in particulars in order to determine generalizations, law deals in generalizations in order to determine particularities.”12 The legal process is supposed to resolve conflicts, often involving the full coercive power of the state, in ways that protect individual rights and in accord with normative community values. The quantitative reasoning of science is not well suited to taking these values or rights into account. In times like the present, when civil rights and community values are under siege from social conservatives, scientific diagnoses in the service of law enforcement and legal adjudication have a particularly strong appeal to those swayed by right-wing ideas. Under these coercive circumstances, the trust in lay judgment, underpinned in part by a particularly American skepticism regarding the authority of experts, comes under fire.
But this acknowledged conflict between science’s truth and the law’s social wisdom assumes that the domain of science is indeed value-free, and sequestered from the social interests of those institutions in government and law and commerce that exercise their authority through the use of scientific knowledge or expertise. A large body of scholarly literature in science studies has challenged this view and has demonstrated that science is no less shaped by social interest than any other field of knowledge. The story about DNA profiling is no exception. It shows how the law enforcement system defines goals for researchers to deliver very particular kinds of knowledge. In a field largely created by the FBI, the direction of DNA forensic research has been wholly governed by the cliental needs of the Justice Department, while the vulnerability of its commercial and police laboratory environments has been fully exposed, and the infallibility of its scientific claims has been hotly contested. Indeed, the meaning of these claims is sufficiently contingent that the probative value of DNA evidence invariably has to be established in the courtroom through lengthy reviews of, or appeals to, the whole peer-review apparatus of scientific and legal literature, as was the case in the O.J. trial. Scientific knowledge, as Sheila Jasanoff concludes in Science at the Bar, is not a simple ancillary to the legal process, waiting to be employed in the pursuit of truth. In many instances, it is highly provisional knowledge, while its authority emerges out of the courtroom battle to prove that the claims of one side’s experts are more contingent than the claims of the other side’s experts.13 People v. Simpson was a dramatic demonstration of this process.
When the probative value of genetic evidence becomes a norm in courtrooms, defendants are more and more likely to be reduced, as sociologists Rochelle Cooper Dreyfuss and Dorothy Nelkin have suggested, to genetic conceptions of “personhood.” Instead of the fully entitled person the legal system is ideally supposed to respect, the defendant is seen as a construct of his or her DNA.14 One of the grave dangers of this tendency is that people brought before the law are once again defined in part by their biological constitution. For centuries, minorities and women have been defined in precisely this way, and more often than not, by loose categories upheld by the law.
In his book White by Law, Ian Haney Lopez has shown that courts historically relied on science and “common knowledge” to determine racial categories, and in particular to determine who counted as white and who did not. These categories used to be governed by the five antique anthropological classifications of Caucasian, Negro, Mongoloid, American (Indian), and Malay. Even so, ethnic and national groups have always been subject to reassignment from one race to another, often in response to changes in immigration laws. By the 1920s, the U.S. Supreme Court’s reliance upon scientific definitions of race had become untenable, and it decided to abandon the use of scientific evidence to adjudicate racial prerequisite cases, and to rely instead upon common sense or popular knowledge. This resulted in the 1922 decision of United States v. Thind, where Bhagat Singh Thind, an Asian Indian, had argued for naturalization on the basis of the Court’s own scientifically refereed equation of “Caucasian” and “white.” Lopez suggests that the Court was increasingly frustrated with science’s inability to accurately identify and quantify racial differences, and that, beginning with Thind, it accepted common knowledge as a more reliable way of policing the boundaries of whiteness, most immediately for the purposes of legislating the 1924 immigration bill.15
The judiciary has continued to play a major role in determining entitlements and benefits along racial lines, and it has actively shaped social beliefs about race in doing so. Overtly biological notions of race were put on the back burner for at least fifty years. In the last two decades, however, these biological definitions have hitched a return ride on the coat-tails of the new molecular genetics. While genetic variation is very slight among humans (and much greater between persons than between population groups), those tiny differences have attained immense cultural significance in a period when biological explanations for social problems are increasingly sought out. The rise of the new genetics has been accompanied by, and, to some extent, utilized in, a brutal rollback of the U.S. state’s affirmative, post—civil rights commitment to closing the racialized gap between the formal equality of its citizens and the material inequalities faced daily by them. Some social science scholarship has made a direct link between the two, none more explicitly than The Bell Curve, whose authors fleshed out their reactionary fantasy of an underclass that was both genetic and cognitive, resistant to all social and educational assistance, and marked by a specific racial profile. More generally, civil libertarians are concerned that the explosion in genetic testing and screening will usher in a social order governed in part by the rule of predictive information about the genetic predisposition of individuals. In this new eugenic dispensation, discrimination is institutionally directed against those genetically designated as risks to society, and thus beyond assistance, along with those profiled as menaces to society, and thus subject to preemptive discipline. This tendency has already been borne out, in the former case, in the realm of medical insurance denied to the “healthy ill,” and in the latter case, in programs like the Violence Prevention Initiative—tried out by the Bush administration—which called for the screening of 100,000 inner-city children to identify potential criminals.
American courts have an appalling record when it comes to safeguarding the rights of minorities. In light of the history, noted above, of the law’s use of science to categorize race, and with an eye to the scary neo-eugenic future of genetically screened population controls, is there any reason for citizens of color to regard the legal application of genetic evidence with anything approaching equanimity? And what bearing, if any, did these concerns have on the O.J. trial?
Despite his initial fame as a “black athlete,” it was widely perceived that O.J. Simpson was less socially black than almost any other black man in America. As comfortable in Brentwood as he was schmoozing with his white friends at the Riviera Country Club in Pacific Palisades or at LAPD Christmas parties, he held down his job as a Hertz spokesperson for an astonishing seventeen years, shilling, in the famous ads, for predominantly white businessmen on tight airport schedules. In an angle already covered in the O.J. joke repertoire (What did Rodney King say to O.J.? “Good thing you didn’t get out of the car, Juice.”), Sistah Souljah declared that the blackest thing O.J. ever did was get chased by the police, even though it was a stretch to call it a chase. Just about the only place where O.J. could not be colorless was in a U.S. court of justice, where no black man has ever felt that the color of his skin has no relevance. More to the point, O.J. was in a court in Los Angeles in the 1990s, facing a charge investigated by law enforcement officers who had established the most racist reputation of any police force since the heyday of white justice in the South. By the end of the trial, many concluded that Simpson had purchased a role for himself in the game of identity politics he had assiduously avoided for so long, and may have done so in order to beat the rap for a gender crime.
From the moment L.A. district attorney Gil Garcetti decided to locate the trial downtown, where the jury pool would be more diverse and where the likelihood of a repeat scenario of the Rodney King uprisings was diminished, the trial had a racialized dimension. With the Fuhrman expose and the defense case built around his racism by Johnnie Cochran (the legal professions crown prince of identity politics), the trial’s focus on race deepened and was amplified a thousandfold in the echo chamber of media reportage. The prevailing media account of public responses to the verdict fiercely reaffirmed all of the divisive fictions about the bichromatic Black and White nation. The most corrosive version suggested that this single, racially weighted, verdict had compensated for centuries of white justice, and that the judicial playing field would be newly leveled as a result. In retrospect, the whole trial became a racial Rosetta stone, endlessly scanned for clues to understanding a society on the brink of abandoning affirmative solutions to its most intractable race problems, while turning a blind eye yet again to the glyphs that spelled out its gender problems.
In all of this drama, little public attention was paid to the single feature of the trial—the data from population genetics used in DNA typing—that formally focused the legal process on O.J.’s racial identity. When the DNA profile obtained from a trace evidence is matched with that of a suspect, statistics derivedfrom population data are applied to estimate the likelihood of a random match. The accuracy of the probability estimate depends on the criteria used for determining the relevantpopulation from which data on allele frequencies can be drawn. In the early years of DNA typing evidence, the reference population was based on the FBI categories of Caucasian, African Americans, and Hispanics, which are far from genetically homogeneous. Dissenting biologists took issue with the treatment of these categories as random mating populations. In 1991, R.C. Lewontin and D.C. Hartl argued that genetic differentiation among ancestralpopulations in Europe is very complex, and immigrant history is too recent and endogamousmating within ethnic groups too high for Caucasian categories to be internally homogeneous. For African Americans, there exists very little reliable data about African slave populations, notwithstanding the extensive mixing with European and Indian ancestry, and the complex genealogical differences between recent Caribbean immigrants and northern urban and southern agrarian blacks. Hispanic is the most vague category, ranging from Guatemalan ancestry that is pure Indian to Argentinian that is pure Caucasian.16 Lewontin, Hartl, and other critics of these categories insisted that significant differences among genetic markers in ethnic subgroups had to be empirically accounted for. Defenders of the process maintained that the degree of variance in population substructures would not substantially differ from profile frequencies calculated from population averages, and that the latter were good enough for estimating random-match probabilities with accuracy.17
The doubts about the subpopulation problem prompted the 1992 NRC commission to recommend a ceiling principle (or more accurately, an interim ceiling principle, in the absence of significant data on ethnic subgroups) based on conservative assessments of data within the major racial groups.18 Since the principle appeared to favor the defendant, there were strong criticisms of this report from scientists and lawyers, and, as is common, questions were raised about conflict of interest: several members of the NRC panel were found to have financial links with companies involved in DNA testing.19 In addition, it was apparent that science in the field was being driven by the urgent needs of the court system and, in particular, by the FBI, which tried to pressure the commission to tone down its reservations on the subpopulation issue. Not surprisingly, the technology’s proponents also tried to clear the field’s name in eminent journals. Lamenting that adversarial lawyers would soon be spreading confusion about science in the upcoming O.J. trial, Eric Lander and the FBI’s Bruce Budowle published their exoneration, “DNA Fingerprinting Dispute Laid to Rest,” in Nature in 1994, only to be further challenged in the journal by Lewontin (who described their article as a “piece of propaganda”) and Hartl, who drew attention to the absurdity of FBI scientists being in a position to determine what constitutes the exact “forensic significance” of statistical differences between racial and ethnic populations.20 Not for the first time in the history of the field, the cops were seen to be guarding the henhouse.
Many of these doubts and controversies were cited by Neufeld and Scheck in People v. Simpson, both in their motion to exclude DNA evidence and in their relentless attempts to cast doubt on the expertise of prosecution witnesses, from lab assistants to leading experts in population genetics. In the meantime, the new NRC commission had been appointed at the request of the FBI director William Sessions in 1993 to stave off a general moratorium, and to settle the disputes about population subdivision and its role in calculating random-match probabilities. The 1996 report declared that the interim ceiling principle was no longer necessary, given the advances in knowledge about the frequency of genetic workers in ethnic subgroups and the overall statistical improvements in genetic profiling of populations. There now existed an “appropriate data base” to support calculations about population substructures, and the data had shown that differences between broad racial groups were more significant than within them.21
The publication of the 1996 report may put to rest some public and institutional doubts about the scientific evidence presented in the O.J. trial, but it resolves none of the issues regarding how complex scientific evidence should be handled in the courtroom nor what uses will be made of the state’s data-banking of DNA samples. For over a century, the state’s criminal agencies have petitioned science with demands for an identifiable self-signature that would serve as a reliable medium of surveillance, classification, and law enforcement. Such signatures, from fingerprints to DNA profiles, have invariably raised, and dashed, the hopes of those looking for biological clues to social behavior. Paul Rabinow points out that the eugenicist Francis Galton was enthused about the first use of fingerprinting in India, motivated by the need of the colonial authorities to counteract the “‘proverbial unveracity’ of the Oriental races,” but was disappointed to find that fingerprints contained no information about race or temperament, no clues about natural or sexual selection.22
The last quarter-century’s research in molecular biology has again attracted the interest of those who believe that people carry within their biological makeup the causes of their own crime, poverty, and poor health. The institutional contexts for DNA forensics in particular have been the laboratories of commercial biotechnology companies, local law enforcement, and the FBI itself, which has its own professionalized research team at Quantico, Virginia, responsible for training forensic specialists in the principles of molecular genetics. These techniques are linked to the FBI’s National DNA Identification Index, developed in coordination with forensic data banks in at least thirty-two states, and based on extensive sampling studies of convicted felons and over two million armed personnel. A number of soldiers, in Hawaii notably, have already refused to be tested, citing the protection of their civil liberties. The DOD, which court-martialed the soldiers, refuses to guarantee that their data banks will not be used for purposes other than identification of the war dead.23 Nonconsensual testing is nonetheless becoming a norm, as Phillip Bereano notes, “at a time of an unprecedented testing hysteria,” where everyone from corporate employees to sex offenders is being subjected to scientific testing.24
The controversy over the courtroom use of DNA typing created a demand for population geneticists to research the differences in DNA marker frequencies among racial and ethnic groups. The FBI alone undertook extensive population surveys resulting in the DNA data-basing of racial and ethnic differences.25 Thus the earlier doubts about population substructure, which were dismissed at the time by the FBI, turned out to be a convenient justification for extending the range of the FBI’s DNA identification banks. The concerns about threats to genetic privacy from this data-banking are considerable. For example, who gets to use the results once the original demand for this kind of research has been satisfied and forgotten? However, there are also larger cultural and social ramifications. The potential outcome of such research will be to reinforce assumptions about the biological classification of ethnicity and race in a social and political climate where such classifications invariably have destructive consequences. Inevitably, it seems, people will once again link cultural and social traits to these biological classifications. Efforts at population control will be able to draw upon biological taxonomies that had proven elusive for centuries. More specifically, genetic testing may be used to affirm the ethnic authenticity of affirmative-action candidates.26 In the world of identity politics, testing may also be used by spokespersons to buttress their claims to be “more black” or “more gay” than others. Anything is possible on this burned-over terrain, which is being cultivated yet again after centuries of overpopulation by destructive scientific and social fictions about race and sexuality.
In the case of DNA typing, there has been no easy separation of the scientific agenda from the social agenda. They have been beset by the same contradiction: Should our identities be analyzed on an individual basis or according to group type? Nothing could be more pivotal to liberalism’s current agony over racial politics. Genetic research has the choice of focusing its resources on the emphasis of individual variation or on group variation. (The former happens to be by far the greater of the two, racial difference accounting for only a small percentage of human genetic diversity.)27 The legal system is obliged to assess subjects as individuals, but it does so within a social climate and within the framework of community and group values where race and ethnicity often have a more powerful significance.
In the O.J. case, this contradiction played an important role. There was arguably less reason to racialize O.J. than with almost any other black American male. While still a black male before the U.S. legal system, he was as “individual” as it was possible to be. And yet, just as the genetic evidence in the trial was based, in part, on references to his racial identity, the social and cultural meaning of his trial was pushed and pulled into the arena of public opinion where race matters—big time. These contradictions between individual and group are nothing new; they lie at the heart of the liberal political system and are increasingly central to public and policy debates about the shape of a multiracial, multiethnic society. But there are many good reasons why appeals to biological authority should be kept to a minimum in such debates. If the criterion of judgment in matters legal, social, and political is whether the genes fit or not, then we are going to be in serious trouble. Consider the outcome in the O.J. trial: the virtual silencing of the prosecution’s establishment of motive in the murder charge, along with much of the supporting evidence of domestic violence. The obsessive focus on DNA identification, which monopolized so much courtroom time, helped to overshadow the systematic evidence of spousal abuse on O.J.’s part. Along with its group-oriented counterpart—the morality play of race politics that Johnnie Cochran’s summation evoked in religion-soaked terms—this drama about identification may well have buried the evidence for the gender crime that many believe lay at the heart of the O.J. case. If so, then the individual/group dualism of the liberal imagination proved yet again that it is not up to the feat of dealing more than two cards at a time.
Thanks are due to Christine Harrington and Dorothy Nelkin for comments and advice on this essay.
1. The California Appellate Court had issued conflicting rulings about the admissibility of the techniques, but was still adhering to the vague criteria of the Frye test for its rulings. The 1923 Washington, D.C, case of Frye v. United States (293 F.2d 1013 D.C. Cir.) had established a precedent for adjudicating the line between “the experimental and the demonstrable stages” of a science, whereby a conservative duration must have lapsed between discovery of a scientific procedure and its forensic application. While the newer techniques of the PCR (polymerase chain reaction) amplification method were at issue, concerns related to the older techniques of RFLP (restriction fragment length polymorphism) analysis were also presented by the O.J. defense team. Evidence based on both techniques was eventually admitted in the trial.
By the time of the O.J. trial, some states, but not California, had adopted the approach favored by the decision in Daubert v. Merrell Dow Pharmaceuticals, Inc. (113 S. Ct. 2786, 1993), which called for a more sophisticated and informed analysis of scientific evidence than Frye had done. The Daubert decision was against Frye’s traditional bias towards peer review and general scientific acceptance, and it reaffirmed the discretionary power of judges to assess the relevance of the scientific evidence to the case in reviewing its admissibility.
2. See Dorothy Nelkin and Susan Lindee, The DNA Mystique: The Gene as a Cultural Icon (New York: Freeman and Co., 1996).
3. Troy Duster, Backdoor to Eugenics (New York: Routledge, 1990), pp. 99—100.
4. 545 N.Y.S. 2d 985 (Sup. Ct. 1989).
5. National Research Council, The Evaluation of DNA Forensic Evidence (Washington, D.C.: National Academy Press, 1996), pp. 1—28.
6. R.C. Lewontin, Biology as Ideology: The Doctrine of DNA (New York: HarperCollins, 1992), pp. 79—82.
7. William F. Buckley, “O.J. on Our Mind,” National Review, vol. 47, no. 12, June 26, 1995, p. 71.
8. Cited in an editorial, Atlanta Journal and Constitution, October 4, 1995, p. 13A.
9. Madam Foreman: A Rush to Judgement? featuring Armanda Cooley, Carrie Bess, and Marsha Rubin-Jackson (Los Angeles: Dove Books, 1995), p. 75. (Cited hereafter in the text as ME)
10. Barry Scheck, “DNA and Daubert,” Cardozo Law Review 15 (1994): 1959–97
11. Lawrence Tribe, “Trial by Mathematics: Precision and Ritual in the Legal Process,” Harvard Law Review 84 (1971): 1329.
12. Marjorie Maguire Schultz, “Reasons for Doubt: Legal Issues in the Use of DNA Identification,” in Paul R. Billings, ed., DNA on Trial: Genetic Identification and Criminal Justice (Plainview, NY: Cold Spring Harbor Laboratory Press, 1992), p. 20.
13. Sheila Jasanoff, Science at the Bar: Law, Science, and Technology in America (Cambridge: Harvard University Press, 1995).
14. Rochelle Cooper Dreyfuss and Dorothy Nelkin, “The Jurisprudence of Genetics,” Vanderbilt Law Review 45, no. 2 (March 1992): 313–48.
15. Ian Haney Lopez, White by Law (New York: NYU Press, 1996), pp. 79—110.
16. R.C. Lewontin and Daniel Hartl, “Population Genetics Problems in the Forensic Use of DNA Profiles,” Science 254, no. 1745 (1991).
17. See Ranajit Chakraborty and Kenneth Kidd, “The Utility of DNA Typing in Forensic Work,” Science 254, no. 1735 (1991); Bruce Weir, “Population Genetics in the Forensic DNA Debate,” Proceedings of the National Academy of Sciences 89, no. 11, 654 (1992); William Thompson, “Evaluating the Admissibility of New Genetic Identification Tests: Lessons from the DNA Wars,” Journal of Criminal Law and Criminology 84, no. 22 (1993): 42—61.
18. National Research Council, DNA Technology in Forensic Science (Washington, D.C: National Academy Press, 1992). Scheck points out that “since the NRC report was published, the overwhelming number of appellate decisions in Frye jurisdictions have rejected methods used by the major forensic laboratories for making statistical estimates, and along with it, the DNA evidence.” Scheck, note 20, p. 1965.
19. Ruth Hubbard and Elijah Wald, Exploding the Gene Myth (Boston: Beacon Press, 1993), p. 151.
20. Eric Lander and Bruce Budowle, “DNA Fingerprinting Dispute Laid to Rest,” Nature 371 (1994): 735; and letters from R.C. Lewontin and Daniel Hartl, Nature 372 (1994): 398—99.
21. In fact, the 1996 report concludes that while “virtually all populations will show some statistically significant departures from random mating proportions … many of the differences will be small enough to be practically unimportant.” NRC (1996), pp. 1–11. The U.S. group that showed the most marked internal variation—a “statistically significant departure”—is American Indian; there do not yet exist adequate databases for many Native peoples’ tribes.
22. Paul Rabinow, “Galton’s Regret: Of Types and Individuals,” in Billings, ed., DNA on Trial, p. 6. See also Making PCR: A Story of Biotechnology (Chicago: University of Chicago Press, 1996), in which Rabinow analyzes the commercial development of the polymerase chain reaction process.
23. The Council for Responsible Genetics and the ACLU have filed an amicus brief in the lawsuit challenging the building of DOD data banks.
24. Philip Bereano, “The Impact of DNA-Based Identification Systems on Civil Liberties,” in Billings, ed., DNA on Trial, p. 121.
25. Federal Bureau of Investigation, VNTR Population Data: A Worldwide Survey (1993), and numerous studies, published by Bruce Budowle et al., in the American Journal of Human Genetics and the Journal of Forensic Science from 1992 to 1995. Even more extensive efforts to classify the genetic range of race have been initiated in the Human Genome Diversity Project, under the auspices of Stanford University.
26. Rabinow, “Galton’s Regret,” p. 17.
27. For all of the advances in DNA data-basing, the degree of human genetic diversity assignable to racial difference has not changed much since Lewontin provided the estimate of 6.3% in 1972. He ended his study with the following proposal: “Human racial classification is of no social value and is positively destructive of social and human relations. Since such racial classification is now seen to be of virtually no genetic or taxonomic significance either, no justification can be offered for its continuance.” “The Approportionment of Human Diversity,” Evolutionary Biology 6 (1972): 397. In the letter in Nature (supra 19), he points out that the FBI’s research efforts in the field of DNA typing would be better devoted to producing “idiotypes,” or unique genetic identifications, rather than the system it has developed, which requires reference to racial groupings.