Jonnie Cochran
Jonnie Cochran; drawing by David Levine


About the case of California v. Simpson, one thing is certain: it has been in no way typical or even illustrative of criminal trials. Simpson’s fame, the huge cast of lawyers, the length of the trial, the intricacy of testimony, the thousands of hours on television, the bizarre taped statements of Mark Fuhrman—there has never been anything like it.

Still, legally speaking, the jury is the only audience. Even as millions have been watching, the evidence and exhibits all aim at swaying the small group who will be voting on a verdict. So for the rest of us, much of the drama consists of trying to guess how members of the panel are reacting to witnesses, reenactments, even to the composure of the man accused of a double murder. To this end, Court TV and CNN call in lawyers as commentators, since they are supposed to know how jurors respond to what happens in a courtroom. However, some recent books on juries, coupled with my own experience, have led me to conclude that lawyers may not be the best judges of how juries and jurors behave. What goes on within those closed rooms, where citizens are essentially on their own, often seems to elude even the shrewdest of legal experts.

Stephen Adler and Jeffrey Abramson provide very different pictures of these private proceedings. Adler discusses five cases, each of which, he seeks to show, resulted in an erroneous verdict. He says they are a fair sample of what is occurring in court-rooms every day. In the case accusing Imelda Marcos of stealing millions from the Philippine government, jurors were manipulated by astute lawyers and ignored convincing evidence showing that she had been stashing money away for years. In another case, where one tobacco company charged another with predatory price-cutting, the jurors understood only a small amount of the testimony. And in two local cases involving murder charges, they believed witnesses of dubious reliability. These results have led Adler to wonder “whether juries do anything well enough to justify our continuing reliance on them.”

Not the least problem, Adler says, is the practice that allows lawyers immense freedom to reject potential jurors. While covering trials as legal editor for The Wall Street Journal, he observed that “many of those who are removed appear to be more alert and unbiased than many who are seated.” Indeed, most of those chosen “are egregiously bad at determining when someone is telling the truth, inadvertently giving false testimony, or lying.” This would happen less often, he writes, if jurors were made aware of research findings showing that “liars typically make fewer hand gestures” and “tend to relax their face muscles.” Judges, he suggests, should “incorporate” such notions “into their legal instructions.”

But to Adler’s mind, even such steps are unlikely to alter the spectacle of juries “failing to see through the cheapest appeals to sympathy or hate.” The system, in his view, tends to winnow out the jurors who might be more skeptical. Jeffrey Abramson provides some figures that may support this view. In a recent year, he tells us, Massachusetts mailed summonses to 905,795 citizens; 253,436 actually appeared, of whom 118,277 were sent to a courtroom and 38,797 were chosen to hear a case. Abramson argues in favor of requiring unanimous verdicts. Currently in most parts of the country, juries in criminal trials are expected to come to a consensus, and, as it happens, they usually do. Nationwide, fewer than 6 percent of criminal cases end with irreconcilable disagreements. The percentage would be further reduced if verdicts of eleven to one, or ten to two, were allowed. But Abramson cites research showing that prosecutors would gain if the numbers of hung juries declined. “The reason,” he writes, is that

among juries deadlocked at 11–1 or 10–2, the holdouts were more than four times more likely to be holding out for acquittal than for conviction. Under a 10–2 rule, therefore, far more split juries return convictions than acquittals.

In describing members of the Simpson jury, commentators tend to emphasize the race and gender of the jurors as well as their age and occupation, the last a surrogate for income and social class. Since the defendant is himself black, the premise of most commentary has been that jurors of his own race will be more likely to accord him a presumption of innocence. For generations, prosecutors could make sure that blacks charged with crimes would be judged by all-white juries and they usually won convictions. Over the last two decades, since its 1975 decision in the case of Taylor v. Louisiana, the Supreme Court has held that jury pools must be drawn from a “cross section” of the community. While that requirement does not ensure that blacks will be selected for particular cases, in practice it has meant that more blacks actually serve. Clearly, race remains a major factor in determining attitudes toward crime. Polls continue to show that most whites who have opinions say they believe Simpson is guilty, whereas most black respondents would vote for acquittal. Hence the widespread view that Simpson’s jurors have biases associated with their race.


The Taylor decision also held that women as well as men must be represented on the lists from which juries are drawn, although so far there has been no requirement that people of different ages and incomes be included. Yet, as with race and gender, these factors, it could be argued, affect not only the ways we evaluate evidence but our overall assessment of the person on trial. Lawyers, sometimes with the aid of consultants, take account of what they suspect may also indicate a juror’s tendencies—loquacity or rigidity, even hairstyles and dress. In the Simpson trial, they had plenty to go on since prospective jurors had to write out answers to 294 questions on their experiences and preferences. Because counsel are allowed dozens of challenges, even unpublicized trials can run through upward of one hundred citizens before finally settling on a panel. As a result, most Americans’ experience with the system has consisted of showing up and being rejected, largely on the basis of lawyers’ intuitions.

Of course the real question is what the twelve supposedly well-screened persons will do when they sit down together. Jeffrey Abramson, after reviewing the relevant research, concluded that “there is no scientific way to predict whether an individual juror will conform, in any one case, to the general attitudes of his or her group.” Michael Knox’s memoir substantiates this view. He served on the Simpson jury from January 11 to March 1, more than enough time to get to know everyone in the sequestered group. Contrary to some news reports, he writes that the jurors did not rigidly divide up according to race or sex or age or class; that both the friendly feelings and the antagonisms among the jurors cut across these lines. As it turned out, the most serious tension was between two black women, both of whom eventually left the jury. Moreover, in transcripts released on July 24, a fifty-year-old black woman juror challenged claims by another black woman juror, who was later dismissed, that the court deputies had favored the whites on the panel. “I don’t think anybody got any more treatment than anybody else,” she said. Indeed, in the confined quarters, personal quirks mattered more than conventional sociological distinctions.

Knox and another former member of the Simpson jury, both of whom are black, have made public their opinions. He says he was “leaning toward a guilty verdict”; while she has called the prosecution case “a whole lot of nothing.”

Except when bias is openly expressed, the elaborate efforts by lawyers and then hired consultants to predict how people will behave if placed on a jury are usually futile. Abramson writes that how they vote will turn mainly on “the fluid group dynamics that influence jury deliberations.” Unfortunately, he says hardly anything beyond this. One reason may be that as a lawyer and a former prosecutor, he has been exempted from jury duty. Adler, who also has a law degree, was once chosen for a civil trial but the case was settled before deliberations began. Both writers spoke with friends who have been jurors; but, oddly, neither book quotes from these conversations. For their information about juries the two writers rely almost entirely on legal scholars, few, if any, of whom have ever been inside a jury room either.

In 1955, a group of University of Chicago professors were allowed, without the jurors’ knowledge, to make recordings of deliberations in a few Kansas civil cases. Word of this got out, an uproar ensued, and over half the states proceeded to outlaw any kind of taping.1 Since then both research projects and real-life litigants have had to depend on simulations. Abramson describes how practice juries have been assembled by lawyers for companies like MCI and General Motors. Various kinds of arguments are tried out before these paid panels, whose deliberations are then observed. These exercises, we are told, help to shape the lawyers’ trial strategies. In academic experiments students are paid to watch films of fictional trials and then deliberate about them. One such project found that requiring unanimity took 138 minutes to reach a verdict, whereas allowing a 10–2 vote cut the time to 103 minutes.2


As it happens I have myself during the past twenty years been on five juries, all of them in criminal cases, including two involving murders. I look forward to being summoned, not as a civic duty but because I find serving an exhilarating experience. So I know what Adler means when he says that speaking with former jurors.


was like having a conversation with someone who was just back from Nepal or who’d just had sex for the first time. They betrayed the same sense of wonder at having been to a new place and having seen life differently.

When before has a citizen been given so grave a responsibility? Freedom or imprisonment; the triumph of truth or of tragic error; adherence to law or malice and caprice—all this and more rests on a juror’s shoulders. Nor will it do to say that the onus is shared with eleven others. In all of the juries I served on, I could see each member acting as if the verdict was his alone. It should be obvious that no simulation can carry this moral weight.

As part of this process, something one could call bonding begins, starting as soon as the last member has been chosen. As anyone who has served on a jury can report, jurors find themselves spending a lot of unsupervised time together, even in proceedings that may last only a week. In the Simpson trial, observers have been struck by how often the jurors have been told to leave the courtroom, sometimes for prolonged periods. In my own experience as a juror, we were usually kept waiting in the morning or after lunch, because the judge had motions from other cases to deal with. Confined to a small anteroom, bus drivers and bond traders are soon on first-name terms. Indeed, I have found it rare for anyone to hide behind a newspaper or a book; in one case we arranged to have lunch together every day at a nearby restaurant. Yet in all our conversations, we, like all other American jurors, were forbidden to talk about the one thing that was uppermost in our minds. The theory is that until the judge’s charge has been given and deliberations begin, only the proceedings themselves should influence the juror’s judgment of the evidence. So during the trial, each juror considers the testimony in isolation, almost as if he were watching the proceedings by himself on television.

What certainly cannot be reproduced by researchers is the process we call deliberation. Like many others, I have served on committees charged with difficult decisions, but none of these matches the bearing and demeanor I have observed in jury rooms. Even is a car-theft case, reaching a verdict took six hours. Perhaps what impresses me most is that no one, at least in my experience, tried to dominate the discussions, and that jurors listened to what the others were saying. After the first half hour, the foreman usually became just another juror.

In my first murder trial, the initial balloting—which we decided to take before we started talking—tied 6–6. Nor did this surprise me. As is usually the case, most of us had in effect been watching different trials. We then sought to move from twelve versions to one, or at least to sufficient consensus for a unanimous verdict. By the fourth hour, three of us changed our votes, now making it 9–4. We were reminded of certain facts we had ignored or felt were unimportant, or we were shown that other interpretations were possible. At one point in writing about the General Will, Jean-Jacques Rousseau said that for democracy to work, citizens must be willing to say, “I was mistaken.” This is not something most of us are willing to do; yet it must happen if juries are to function, and I have observed more than a few of my fellow New Yorkers make that admission. About two hours later, we were down to 10–2, and it then took another four to arrive at 12–0. At no point were the divisions along racial or other social lines.

And as we moved toward unanimity, it became evident that each of us was gaining a broader perspective. That twelve minds can take in more than one is the epistemological rationale for the jury system. But the rationale only works when bond traders are willing to listen to bus drivers and learn from them. Thus far the only place I have found that happening is inside the jury room.

In theory, juries attend to the facts, leaving the judge to settle matters of law. If only the distinction were that simple. Hazel Thornton’s account of the Menendez trial shows that the “facts” her jury had to find were actually legal constructions. As everyone must know, Erik and Lyle Menendez admitted that they had killed their parents, claiming that there were exonerating circumstances, ranging from preemptive self-defense against sexual abuse, and even the threat of murder by their father, to impaired mental capacity.3 They stressed that they were not confessing to a crime. For its part, the prosecution argued that a felony had been committed and the brothers must be punished.

So while the jury was relieved from finding whether the sons had caused their parents’ deaths, they were given what turned out to be an even more formidable task, deciding what class of offense—if any—had been committed. As will be recalled, two juries sat in the courtroom, one for each brother. Thornton was assigned to the jury hearing the case of Erik, the younger brother. To complicate matters further, she writes, each jury had to decide whether its defendant had committed the same order of crime when killing the two parents or if the two deaths represented two types of slayings.

Nor would things get simpler. The prosecution gave each jury the same set of choices. Each brother was charged with having committed all four of the following crimes against each of his parents: first-degree murder, which carries a mandatory death sentence in California; second-degree murder; voluntary manslaughter; involuntary manslaughter. And, as the defense said, there was a fifth alternative: that they had committed no crime at all. In law and in logic, the young men could not have committed all four or even any two of the offenses that had been entered against them. So the jury’s job was to determine which offense—if any—took place when each defendant participated in each death. Thus Erik’s shooting his father might be defined as second-degree murder, while Lyle’s firing at his mother could be involuntary manslaughter—or vice versa. (Votes would also be taken on charges of conspiracy.)

After almost five months of testimony, the juries retired to deliberate. Some crime had clearly been committed, but which of the four it was for each brother depended on various combinations of conditions. Thus one of the “facts” each jury had to ferret out was whether the particular brother they were concerned with had “intended” to kill either or both of his parents. (Needless to say, each jury also spent a lot of time speculating about the other brother, since either brother’s intention could turn on the relations he had with the other.) At issue too was whether “intent” to kill, which requires a clear aim to bring about death, might have been accompanied by “premeditation,” i.e., a previously planned scheme to kill, although the distinction between the two is often murky. In first-degree murder, both states of mind must be present, and they must also be combined with “express malice,” i.e., a clearly indicated desire to harm which, again, is all too easily confused with the other legally required states of mind. Second-degree murder also needs intent, but does not require premeditation; the jury had to find only “implied malice.”

The two different manslaughter charges contained similarly puzzling distinctions, hinging on how far the accused were in control of themselves and to what degree, if any, they had themselves been abused or had suffered psychologically from sexual abuse. And each jury was expected to reach a 12–0 agreement—including agreement about motives—that “factually” described each brother’s shooting of each parent.4

Not surprisingly, Thornton tells us that her jury kept going back to the judge for clarifications on definitions, and each time he read them abstractly phrased standard definitions that offered little enlightenment.5 After 106 hours of deliberations, she and the others on Erik’s jury declared they could not reach a unanimous decision; and thirty-three hours later, Lyle’s did the same. What most of us recall is that the trial ended with a “hung jury.” Yet as Thornton’s book makes clear, this does not adequately describe what happened. All twenty-four of the jurors felt that the brothers had committed a crime when they killed both of their parents. But they divided sharply over which of the charges best described what had occurred. Only one of the twenty-four felt that involuntary manslaughter was an appropriate verdict, and that juror applied it only to Lyle’s slaying of his mother.

The gender of the jurors counted heavily. In Erik’s case, six men, impressed by the brothers’ apparent interest in inheriting money, found him guilty of murdering his parents, while six women thought the intimations that he had been sexually abused by his father raised enough doubt about his motives to justify a finding of manslaughter. Mrs. Thornton’s diary entry for the day the jury heard one of the most important pieces of evidence is revealing of her own state of mind:

Okay now we’ve heard “THE TAPE,” the real “confession” tape in which Dr. Oziel recorded part of an actual session with the boys, with their cooperation, on 12/11/89. This is supposedly the most damaging evidence against them but, in my own mind, the way things have been developed over the past four months, it only served to strengthen their defense! True, they never directly mentioned self defense or abuse on the tape, but it could be construed (and was construed by Ms. Abramson and Dr. Burgess) that there were many allusions to both.

Clearly the defense lawyer, Leslie Abramson, made a strong impression on Ms. Thornton and her fellow women jurors. (“All I know is, win or lose, she is worth the one million dollars I heard she was being paid.”) But the bitter split between men and women in this case seems to be exceptional. Jeffrey Abramson notes that women jurors tend to talk less than men and are more apt to vote for guilt in rape cases. Even so, in general, he finds “no evidence that women and men end up voting as blocs on juries.”

As matters now stand, jurors are not forewarned that they may have to diagnose mental and emotional conditions like premeditation, intention, and malice. Nor, at least in some cases, should they be asked to. The table below shows how what was being demanded of the Menendez case juries was virtually impossible:


to come to four different 12–0 decisions on a tangled web of words far removed from any recognizable realities.

Mrs. Thornton hardly mentions the mandatory death penalty but it is likely it was very much on the minds of the jurors, particularly the women jurors, and that it was not objectionable to six of the male jurors. Indeed, as the deliberations open, Mrs. Thornton writes, “my worst nightmare has come to pass. Our initial vote, specifically as to the charge of Murder One of Jose [the father]: Six Men—Guilty, Six Women—Not Guilty.”

What happened in the Menendez case supports Stephen Adler’s argument that we can ask too much of jurors. Why not let them decide whether the defendant has knowingly committed a more simply defined felonious act—instructions on that could be given—and then give the judge broader scope to determine the class and severity of the offense? Such a reform would have to be very carefully worked out, and it could allow jurors to exclude the death penalty. It would also avoid much of the needlessly bewildering complexity that juries must now contend with. In fact, something rather like this happens when, after hearing the jury’s verdict, the judge sets the length of the sentence, although the discretion allowed federal and state judges is unfortunately more and more limited.6


In one of the trials I observed as a juror we, too, had to reach a verdict not simply by establishing the facts but by defining a legal conception. A young man was charged with “resisting arrest.” The facts, as we later reconstructed them from the testimony of two policemen and three witnesses, were that an officer had told him to stop but he had continued walking. One of the officers tackled him, put on handcuffs, and hustled him into a patrol car. In our deliberations, we agreed early on that the defendant had ignored the officer.

But was the defendant walking away from an “arrest”? None of the witnesses—including the policemen themselves—testified that the officer had said, “You are under arrest.” No one claimed that the man had been orally charged with a specific crime, or that he had been read his rights. When we took our questions to the judge, she told us that the law did not have a precise definition of “arrest,” so it was for the jury to decide whether one had actually occurred.

Although this jury had on it a Pulitzer prize-winning columnist known for his commentaries on urban life, the best answer came from a hospital kitchen worker, who pointed out that it would be impossible for police to function at all if civilians could walk away with impunity when told to stop. In fact, etymology prevailed: if an officer chooses to stop you, you are then and there “arrested.” This is so even if he simply signals you, without saying a word. So being under arrest requires no formalities and might last less than a minute. At least so we concluded, and the judge said she liked our definition. Since all this occurred in a city court, it seems unlikely that our definition will be adopted in case law.

While I continue to have doubts about presenting jurors with several charges for a single offense, I should add in fairness that such options sometimes serve a practical purpose: graded alternatives give jurors room to maneuver. In one of my trials, a young woman plying her profession was being harassed by a drug dealer. She called for her protector, who dashed gun in hand from a nearby bar and proceeded to fire. As it happened, he missed his target and killed a bystander. We all managed to agree that a slaying had been “intended,” which made it some kind of murder. But two of us were not persuaded that it had also been “premeditated,” which was necessary for a first-degree conviction. So the rest of us offered to vote for second-degree murder to get a unanimous verdict. Since we knew that a lesser charge would bring a lighter sentence, we were, in effect, engaging in a variant of plea-bargaining on behalf of the defendant. Legally, this should not happen and judges warn against it, since sentencing is supposed to be outside the jury’s responsibility. Still, everyone knows there would be far fewer 12–0 votes if juries adhered to the letter of the law.

Throughout the Simpson trial, there has been only one charge: first-degree murder, which was all the overconfident prosecution chose to file. This, as has been noted, requires that all the jurors agree that there was intent, premeditation, and malice. However, California law allows the judge, at a trial’s end, to give the jury more choices. The defense may express an opinion to the judge on this matter, and it would probably press for the single, original charge as the best way of getting a divided jury. While the prosecution can no longer seek a longer list of charges, they would probably now prefer one in hopes of salvaging some kind of conviction.7

Michael Knox devotes most of his book to describing how sequestration has affected the Simpson jurors. It is a dispiriting story. How those who remain have endured this confinement for more than seven months defies comprehension. While each has a single room on a special hotel floor, they are not allowed to visit one another even to exchange a few words. The usual, and entirely defensible, prohibition against discussing the case has been carried to extreme lengths, reflecting a degree of caution and mistrust that seems to me absurd. Most personal conversations are prohibited. The only choices are to be alone in your own room or to be thrown in with all the others, where the main activities are watching television or chatting warily in everyone’s else’s hearing. Armed and uniformed deputies are always on hand. Nor have these adults been allowed even a glass of wine at dinner or a few beers at the weekend. For a couple of jurors to have dinner together in one of their rooms is out of the question. All these restrictions are based on the assumption that if the jurors are left alone for an instant outside the sight of a deputy, they will start discussing the trial. The same stricture holds with visitors, who can only be seen in the common room, with exceptions made for spouses who, according to a dismissed juror, can make conjugal visits only between 7 PM and midnight and can’t leave earlier even if they want.

They live in a state of collective solitary confinement, and some resent it. “When you’re sequestered,” Knox observes, “you can’t avoid anyone.” And, as in all confined settings, “minor incidents became major blow-ups,” exacerbating personal enmities. Some liked to chat at breakfast, while others preferred silence. Nor would separate tables solve the problem, since they shared a small dining room. Reading this, I see that the jury bonding I mentioned earlier worked when all of us went home at the end of each day, and wonder whether the many months of confinement in a Los Angeles hotel will affect the trial’s outcome. Some have predicted that the jurors will blame the state and the prosecution for unreasonable restrictions. Perhaps the discipline of deliberations will overcome the animosities. Still, tensions may recur in the evenings and over weekends, where they will again be under guard and at which time discussing the case will again be forbidden.

All of which raises the question of whether sequestration is necessary at all, even in highly publicized cases. My original view was that if jurors were allowed to go home, they might have to fend off friends or strangers who wanted to discuss the case, not to mention intimidating mail or telephone calls, since there would be no assurance that their identities would remain secret. But I learned from Hazel Thornton’s book, Hung Jury, that the Menendez jurors were permitted to return home every evening, not only while testimony was being heard but also during deliberations. In the publicity and popular feeling it aroused, the Menendez trial was not far different from the Simpson case, yet apparently no juror was approached. Nor is there reason to believe that any of them read or watched reports of the trial. They failed to reach a verdict for reasons I have mentioned and not because leading a more normal life put their objectivity in doubt.

And what of Stephen Adler’s proposal that we jettison juries altogether? With complex civil trials, he has a strong case. In such matters as patent infringement, securities fraud, anti-trust violations, even product liability, decisions could well be left to a judge. However with charges such as libel and employment discrimination, the facts are more readily grasped and deserve a public forum. For criminal cases, neither Adler nor other critics have shown that judges acting alone are more likely to render convincing verdicts. The issue goes beyond deciding whether an estranged husband committed murder outside a house in Brentwood. And here Viscount Hewart’s stricture still holds: “Justice should not only be done, but should manifestly and undoubtedly be seen to be done.” Requiring unanimity from a panel of twelve people goes a long way toward assuring the rest of us that, whether doubts prevailed or were overcome, the verdict was based on careful and often painful thought.

August 24, 1995

This Issue

September 21, 1995