University of Miami Law School University of Miami School of Law Institutional Repository

Articles Faculty and Deans

2003 Trials and Tribulations: Science in the Law Susan Haack University of Miami School of Law, [email protected]

Follow this and additional works at: https://repository.law.miami.edu/fac_articles Part of the Evidence Commons, and the Law and Society Commons

Recommended Citation Susan Haack, Trials and Tribulations: Science in the Law, Fall Daedalus 54 (2003).

This Article is brought to you for free and open access by the Faculty and Deans at University of Miami School of Law Institutional Repository. It has been accepted for inclusion in Articles by an authorized administrator of University of Miami School of Law Institutional Repository. For more information, please contact [email protected]. Susan Haack

Trials & tribulations: science in the courts

“I should like to know” [asked Mr. Chi- mining factual , in both criminal chely] “how a coroner is to judge of evi- and civil cases, courts very often need dence if he has not had a legal training?” to call on scientists: on toxicologists and “In my opinion,” said Lydgate, “legal tool-mark examiners, epidemiologists training only makes a man more incompe- and engineers, serologists and psychia- tent in questions that require trists, experts on pcbs and experts on of another kind. People talk about evi- paternity, experts on rape trauma syn- dence as if it could really be weighed in drome and experts on respiratory disor- scales by a blind Justice. No man can judge ders, experts on blood, on bugs, on bul- what is good evidence on any particular lets, on battered women, etc. For, as sci- subject unless he knows that subject well. ence has grown, so too has the legal A lawyer is no better than an old woman system’s dependence on scienti½c evi- at a post-mortem examination. How is he dence; it has been estimated that by 1990 to know the action of a poison? You might around 70 percent of cases in the United as well say that scanning verse will teach States involved expert testimony, most you to scan the potato crops.” of it scienti½c. Such testimony can be a powerful tool for justice; but it can also –George Eliot, Middlemarch (1872) be a powerful source of confusion–not to mention opportunities for oppor- Justice requires just laws, of course, and tunism. just administration of those laws; but it Who could have imagined, when dna also requires factual truth. And in deter- was ½rst identi½ed as the genetic materi- al half a century ago, that dna analysis Susan Haack is Cooper Senior Scholar in Arts and would by now have come to play so large Sciences, professor of philosophy, and professor of a role in the criminal justice system, and law at the University of Miami. Internationally in the public of the law? known for her work in philosophy of , episte- Even twenty years ago, forensic scien- mology, , , and tists could tell only whether a blood the law of scienti½c testimony, Haack is the author sample was animal or human, male or of several books, most recently “Manifesto of a female, and, if human, of what type (the Passionate Moderate” (1998) and “Defending least common blood type being found in Science–Within Reason: Between and 3 percent, and the commonest in 43 per- Cynicism” (2003). cent, of the U.S. population). Then, in the mid-1980s, dna ‘½ngerprinting’ © 2003 by Susan Haack made vastly more accurate identi½cation 54 Dædalus Fall 2003 possible, to probabilities of the order of on the reliability of memory, perception, Science in a billion to one; and by now new tech- and eyewitness testimony, that less the courts niques have made it possible to amplify than half a century later psychological and test the tiniest samples. evidence would play a signi½cant role At ½rst, such evidence was strenuously in such landmark constitutional cases contested in court; but as its solidity, as Brown v. Board of Education (1954), or and its power to enable justice, became that by now it would have come to play unmistakable, the ‘dna wars’ gradually so large a role in the criminal justice sys- died down. By the spring of 2002, dna tem–or that it would be the focus of testing had exonerated more than a hun- seemingly endless controversy? For dred prisoners, including a signi½cant while the work of experimental psy- number on death row, and helped con- chologists on eyewitnesses, memory, vict numerous rapists and murderers. In etc., has indeed proved useful, clinical at least one instance, it both exonerated psychologists’ and psychiatrists’ diag- and convicted the same person: after noses of this syndrome and that, and es- serving nearly eleven years of a twenty- pecially their theories about the repres- ½ve-to-½fty-year sentence for rape, Kerry sion and recovery of traumatic memo- Kotler was released in 1992 when newly ries, have been the subject of heated bat- conducted dna tests established his in- tles in the courtroom, in the press, and nocence; less than three years after his in the academy. release, he was charged with another In the mid-1980s, testimony of alleg- rape, and this time convicted on the ba- edly repressed and recovered mem- sis of dna analysis identifying him as ories came to public attention in the the perpetrator. McMartin Preschool case–the longest Even so, dna evidence can present U.S. criminal trial ever (six years), and problems of its own: police of½cers and one of the most expensive (around $15 forensic technicians make mistakes– million). But in 1990 the seven defen- and have been known deliberately to fal- dants were acquitted of the ritual sexual sify or misrepresent evidence; juries may abuse that, under the influence of thera- misconstrue the signi½cance of expert pists, numerous children at the school testimony about the probability of a ran- had claimed to remember. George dom match with the defendant, or of in- Franklin spent nearly seven years in formation about the likelihood that a prison for the murder of nine-year-old sample was mishandled–and attorneys Susan Nason, convicted on his daugh- have been known to contribute to such ter’s supposed memory of the event, misunderstandings; criminals devise recovered under hypnosis twenty years devious ways to circumvent dna iden- afterward; he was released in 1996, after ti½cation–and at least one prisoner, ap- his daughter also ‘remembered’ his com- parently hoping to exploit the potential mitting two other murders, with respect for confusion, has petitioned for a dna to one of which he could be unambigu- test that, as he must have anticipated, ously ruled out. (Franklin later sued con½rmed his guilt. prosecutors and the experts who testi- And who could have imagined, when ½ed against him for wrongful prosecu- Hugo Münsterberg urged in his On the tion and violation of his civil rights.) By Witness Stand: Essays on Psychology and the late 1990s, it began to seem that crit- Crime (1908) that the law avail itself of ics such as experimental psychologist the work of experimental psychologists Elizabeth Loftus, who had maintained

Dædalus Fall 2003 55 Susan all along that supposedly repressed and atric and clinical testimony, experimen- Haack on recovered ‘memories’ could be the result tal psychologist Margaret Hagen writes science of therapists’ suggestive questioning, of “charlatans and greedy frauds.” were vindicated. But recently the ‘mem- But other scientists–like Eliot’s Dr. ory wars’ have flared up all over again, Lydgate–think the real problem is, rath- this time in legal claims ½led against er, that jurors, attorneys, and judges are Catholic priests accused of sexual abuse too illiterate scienti½cally to discrimi- of children and young people. nate sound science from charlatanism. Norman Levitt, for example, comment- Why has the legal system found scien- ing in Prometheus Bedeviled (1999) on the ti½c testimony hard to handle? Ever “noisome travesty” of the O. J. Simpson since there have been scienti½c witness- trial, complains that “the basic princi- es, lawyers and legal scholars–like ples of statistical inference were opaque Eliot’s Mr. Chichely–have had their to all concerned except the witnesses doubts about them. The commonest themselves. The lawyers . . . , the judge, complaint has been that venal scientists the dozens of commentators . . . , and brought in by unscrupulous attorneys certainly the woozy public–all seemed will testify to just about anything a case utterly ignorant as to what . . . statistical demands. In 1858, the Supreme Court independence might mean . . . . All the observed that “experience has shown other scienti½c issues encountered the that the opposite opinions of persons same combination of neglect and eva- professing to be experts may be ob- sion.” tained in any amount”; in 1874, John Or- There surely are venal and incompe- dronaux wrote in the American Journal of tent scienti½c witnesses, and there surely Insanity that “If Science, for a consider- are scienti½cally ignorant and credulous ation, can be induced to prove anything jurors, attorneys, and judges; but the fa- which a litigant needs in order to sustain miliar complaints gloss over many com- his side of an issue, then Science is fairly plexities. Scienti½c testimony may be open to the charge of venality and per- flawed by outright fraud, or, more often, jury, rendered the more base by the dis- by the overemphatic presentation of guise of natural truth in which she robes scanty or weak evidence; it may be solid herself.” More than a century later, in science misapplied by a poorly run labo- Galileo’s Revenge (1991), Peter Huber was ratory, or serious but highly speculative sounding a similar theme: junk science and controversial science, or sloppily –“data dredging, wishful thinking, truc- conducted scienti½c work, or pseudo- ulent dogmatism, and, now and again, scienti½c mumbo jumbo. The motive outright fraud”–was flooding the may be an expert’s greed, or his desire to courts. Some scientists concur. In her feel important, or his anxiety to help the study of the silicone breast implant ½as- police or a sympathetic plaintiff; or it co, Science on Trial (1996), Marcia Angell may be a scientist’s conservatism about complains that “[e]xpert witnesses may new and radical-sounding ideas; or a wear white coats, be called ‘doctor,’ pur- plaintiff’s attorney’s interest in keeping port to do research, and talk scienti½c disputes long settled in science legally jargon. But too often they are merely alive. Failures of understanding may be adding a veneer to a foregone, self- due to jurors’ or judges’ or attorneys’ interested conclusion”; in Whores of the inability to follow complex statistical Court (1997), an exposé of flimsy psychi- reasoning, or to their ignorance of the

56 Dædalus Fall 2003 kind of controls needed in this or that ½nality. But these pragmatic and piece- Science in type of experiment or study, or to their meal strategies, though in some ways the courts excessive deference to science, or their more promising, raise hard questions resentment of its perceived elitism. Or about why we value trial by jury, why we the problem may simply be jurors’ sense want ½nality, and whether the adversar- that someone should compensate the vic- ial process is really an optimal way of tim of an awful disease or injury, or that ensuring–in the words of the preamble someone should be punished for a horri- to the Federal Rules of Evidence–“that ble crime. the truth be ascertained.” And the familiar complaints also gloss over the deep tensions between science The present practice of relying on ex- and the law that are at the root of these perts proffered by the parties not to re- problems. The culture of the law is ad- port on what they saw but rather to give versarial, and its goal is case-speci½c, ½- their informed opinion, evolved only nal answers. The culture of the sciences, gradually, along with the growth of the by contrast, is investigative, speculative, adversary system, cross-examination, generalizing, and thoroughly fallibilist: and formal rules governing the admissi- most scienti½c conjectures are sooner or bility of evidence. For a long time it was later discarded, even the best-warranted required only that a scienti½c witness, claims are subject to revision if new evi- like any other expert witness, establish dence demands it, and progress is ragged his quali½cations as an expert–until and uneven. Science doesn’t always have 1923, when the Frye1 ruling imposed new the ½nal answers the law wants, or not restrictions on the proffered testimony when it wants them; and even when sci- itself. ence has the answers, the adversarial In Frye, excluding testimony of a then process can seriously impede or distort new blood-pressure deception test, the communication. It’s no wonder that the D.C. court ruled that novel scienti½c evi- legal system often asks more of science dence was admissible only if it had than science can give, and often gets less gained “general acceptance in the ½eld from science than science could give; to which it belongs.” At ½rst cited only nor that strong scienti½c evidence some- quite rarely, and almost always with re- times falls on deaf legal ears, while flim- gard to lie-detector evidence, the Frye sy scienti½c ideas sometimes become le- rule gradually came to be widely fol- gally entrenched. lowed in criminal trials, and by 1979 had One response to the dif½culties has been adopted in a majority of states. (It been to try to tame scienti½c testimony remains of½cially the law today in a by devising legal rules of admissibility to number of states, Florida included.) Of ensure that judges don’t allow flimsy course, general acceptance is a better stuff to be presented to juries. But, as the proxy for scienti½c robustness when the tortuous history of efforts to frame such ½eld in question is a mature, established formal rules suggests, no legal form of scienti½c specialty than when it is a words could guarantee that only good- highly speculative area of research–or, enough scienti½c testimony is admitted. worse, the professional turf of a trade Another response has been, instead, to union of mutually supportive charlatans. adapt the culture of the law, bringing it Moreover, the rule is highly manipula- more into line with science by compro- 1 Frye v. United States, 54 App. D.C. 46, 293 mising adversarialism or the concern for F. 1013 (1923).

Dædalus Fall 2003 57 Susan ble, depending, among other things, on scandalously weak scienti½c testimony Haack on how broadly or narrowly a court con- that would have been excluded under science strues the ½eld in question. Neverthe- Frye but was being admitted under the less, a main focus of criticism was that Federal Rules. Then in 1993, with pro- the Frye test was too restrictive. posals before Congress to tighten up The Federal Rules of Evidence (1975) the Federal Rules, the Supreme Court is- seemed to set a less restrictive standard: sued its ruling in the landmark Daubert the testimony of a quali½ed expert is ad- case3–the ½rst case in the Court’s 204- missible provided only that it is relevant, year history where the central issue was and not legally excluded on grounds of the standard of admissibility of scienti½c unfair prejudice, waste of time, or po- testimony. tential to confuse or mislead the jury. In Daubert was a tort action against line with the Federal Rules’ apparently Merrell Dow Pharmaceuticals brought liberal approach, in Barefoot,2 a 1983 by parents who claimed that their chil- constitutional case, the Supreme Court dren’s severe birth defects had been af½rmed that the rights of a Texas defen- caused by their mothers’ taking the dant were not violated by the jury’s company’s morning sickness drug, Ben- being allowed, in the sentencing phase, dectin, during pregnancy. In excluding to hear psychiatric testimony predicting the plaintiffs’ expert testimony, the his future dangerousness–even though lower court had cited Frye (which up an amicus brief ½led by the American till then, contrary to Huber’s diagnosis, Psychiatric Association reported that had almost always been cited in crimi- two out of three psychiatric predictions nal, not civil, cases). Remanding the of future dangerousness are mistaken. case, the Supreme Court held that the Justice White, writing for the majority, Federal Rules had superseded Frye, but observed that the Federal Rules antici- added that the Rules themselves re- pate that courts will admit relevant evi- quired judges to screen proffered expert dence and leave it to juries, with the help testimony not only for relevance, but al- of cross-examination and presentation so for reliability. of contrary witnesses, to determine its Justice Blackmun wrote for the majori- weight. In dissent, however, noting that ty that courts must look not to an ex- a scienti½c witness has a special aura of pert’s conclusions, but to his methodol- credibility, Justice Blackmun averred ogy, to determine whether proffered tes- that “[i]t is extremely unlikely that the timony is really “scienti½c . . . knowl- adversary process will cut through the edge,” and hence reliable. Citing law facade of superior knowledge.” professor Michael Green citing philoso- By the late 1980s, as legal scholars de- pher of science , and adding bated whether the Federal Rules had or a quotation from Carl Hempel for good hadn’t superseded Frye, and whether a measure, the ruling suggested four fac- more or a less restrictive approach to sci- tors for courts to consider: falsi½ability, enti½c testimony was preferable, there i.e., whether the proffered evidence can was rising public and political concern be, and has been, tested; the known or that the tort system was getting out of potential error rate; peer review and hand; a crisis due in large measure, publication; and (in a nod to Frye) ac- Huber argued in his influential book, to ceptance in the relevant scienti½c com- 2 Barefoot v. Estelle, 463 U.S. 880, 103 S.Ct. 3383 3 Daubert v. Merrell Dow Pharm. Inc., 509 U.S. 579, (1983). 113 S.Ct. 2786 (1993).

58 Dædalus Fall 2003 munity. Dissenting in part, however, Jus- But Popper’s and Hempel’s philoso- Science in tice Rehnquist pointed out that the word phies of science are not compatible. the courts ‘reliable’ nowhere occurs in the text of Worse, neither can supply the hoped-for Rule 702; anticipated that there would crisp criterion to discriminate the scien- be dif½culties over whether and how ti½c, and hence reliable, from the un- Daubert should be applied to nonscien- scienti½c, and hence unreliable. No phi- ti½c expert testimony; worried aloud losophy of science could do this; no such that federal judges were being asked to criterion is possible, for not all scientists, be amateur scientists; and questioned and not only scientists, are good, reliable the wisdom of his colleagues’ foray into inquirers. Nor is there a uniquely ratio- the philosophy of science. nal mode of inference or procedure of That foray was indeed (if you’ll pardon used by all scientists and only by the expression) ill judged. As Justice scientists–no ‘scienti½c method’ in the Blackmun’s ellipses acknowledge, Rule sense the Court assumed. Rather, as Ein- 702 doesn’t speak of “scienti½c knowl- stein once put it, scienti½c inquiry is “a edge,” but of “scienti½c or other techni- re½nement of our everyday thinking,” cal knowledge.” However, doubtless in- superimposing on the inferences, desid- fluenced by the honori½c use of “sci- erata, and constraints common to all ence” and “scienti½c” as all-purpose serious empirical inquiry a vast variety terms of epistemic praise, the majority of ampli½cations and re½nements of hu- apparently took for granted that there is man cognitive powers: instruments of some mode of inference or procedure of observation, models and metaphors, inquiry, some methodology, that is dis- mathematical and statistical techniques, tinctive of genuinely scienti½c, and experimental controls, etc., devised by hence reliable, investigation. And so generation upon generation of scientists, they reached for Popper’s criterion of constantly evolving, and often local to demarcation, according to which the this or that area of science. hallmark of genuine science is that it is So perhaps it is no wonder that in the falsi½able, i.e., could be shown to be two subsequent decisions in which it has false if it is false; and for his account of spoken on the admissibility of expert the scienti½c method as conjecture and testimony, the Supreme Court quietly refutation, i.e., as making bold hypothe- backed away from the confused philoso- ses, testing them as severely as possible, phy of science built into Daubert. In the and, if they are falsi½ed, giving them up Court’s ruling in Joiner4 (a toxic tort case and starting again rather than protecting involving pcb exposure), references to them by ad hoc maneuvers. Unfortu- Hopper, Pempel, falsi½ability, scienti½c nately, however, Popper’s philosophy of method, etc., are conspicuous by their science is singularly ill suited as a guide absence; and the distinction between to reliability; for, if he were right, sci- methodology and conclusions, crucial to enti½c theories could never be shown Daubert, is repudiated as not really viable to be true or even probable, but at best after all. And in response to inconsistent “corroborated,” by which Popper means rulings across the circuits over the appli- only “tested but not yet falsi½ed.” And cability of Daubert to nonscienti½c ex- so the Court ran Popper together with perts, in Kumho5 (a product liability case Hempel, whose logic of con½rmation 4 General Electric Co. v. Joiner, 522 U.S. 136, 118 does allow that scienti½c claims can be S.Ct. 512 (1997). con½rmed as well as discon½rmed. 5 Kumho Tire Co. v. Carmichael, 526 U.S. 137, 119 S.Ct. 1167 (1999).

Dædalus Fall 2003 59 Susan involving a tire blowout) the Court ruled from weak, the Daubert ruling involved Haack on that Daubert applies to all expert testi- a signi½cant shift of responsibility from science mony, not only the scienti½c. According juries to judges, a shift Justice White to the Kumho Court, the key word in Rule had resisted. As Judge Alex Kozinski, to 702 is “knowledge,” not “scienti½c”; whom Daubert was remanded,6 causti- what matters is whether proffered testi- cally observed, he and his colleagues mony is reliable, not whether it is sci- “face a far more complex and daunting ence. task in a post-Daubert world . . . . [T]hough However, the Supreme Court certainly we are largely untrained in science and didn’t back away from its commitment certainly no match for any of the wit- to federal judges’ gatekeeping responsi- nesses whose testimony we are review- bilities. Far from it. In Joiner, the Court ing, it is our responsibility to determine af½rmed that a judge’s decision to allow whether the experts’ proposed testimo- or exclude scienti½c testimony, even ny amounts to ‘scienti½c knowledge,’ though it may determine the outcome of constitutes ‘good science,’ and was de- a case, is subject only to review for abuse rived by the ‘scienti½c method.’” In a of discretion, not to any more stringent post-Kumho world, the task is even more standard. And in Kumho, stressing that daunting. the factors listed in Daubert are “flexi- In the wry words of Federal Judge ble,” the Court ruled that a judge may Avern Cohn: “You do the best you can.” use any, all, or none of them. So, aban- A sensible layperson might suspect that doning the false hope of ½nding a form an expert witness is confused, self- of words to discriminate “reliable, scien- deceived, or dishonest, or that he has ti½c” testimony from the rest, the Kumho failed to take account of readily available Court left federal judges with wide-rang- relevant information; and should be ing responsibility and considerable dis- capable of grasping the importance of cretion in determining whether expert double-blinding, independence of vari- testimony is reliable enough for juries to ables, etc. But the fact is that serious ap- hear, but with little guidance about how praisal of the worth of complex scien- to do this. ti½c evidence (as Dr. Lydgate pointed Though the Daubert ruling spoke of out long ago) almost always requires the Federal Rules’ “preference for ad- much more than an intelligent layper- missibility,” it imposed signi½cantly son’s understanding of science: the spe- more stringent requirements than Jus- cialized knowledge needed to realize tice White had envisaged in Barefoot; that an experimenter failed to control arguably, indeed, more stringent re- for this subtle potentially interfering quirements than Frye. (In 2000, revised factor; that these statistical inferences Federal Rules made explicit what, ac- failed to take account of that subtle de- cording to Daubert, had been implicit in pendence of variables; that new work Rule 702 all along: admissible expert tes- has cast doubt on this widely accepted timony must be based on “suf½cient” theory; that this journal is credible, that facts or data and be the product of “reli- journal notorious for such-and-such edi- able” principles or methods, which the torial bias. witness has “reliably” applied to the Since Daubert there have been various facts of the case.) And, despite the usual efforts to educate judges in science– rhetoric about the Court’s con½dence such as the two-day seminar on dna for in the adversarial system and in jurors’ 6 Daubert v. Merrell Dow Pharm. Inc., 43 F.3d 1311 ability to sift strong scienti½c testimony (1995). 60 Dædalus Fall 2003 Massachusetts Superior Court judges at When the report was made public, a Science in the Whitehead Institute for Biomedical headline in The Washington Post hailed it the courts Research, after which, the director of the as a “Benchmark Victory for Sound Sci- institute told The New York Times, they ence,” and an editorial in The Wall Street would “understand what is black and Journal announced that “reason and evi- white . . . what to allow in the court- dence have ½nally won out.” And it is room.” But while a bit of scienti½c edu- not only those whose sympathies lie cation for judges is certainly all to the with defendant companies in danger of good, a few hours in a science seminar being bankrupted by baseless tort claims will no more turn judges into scientists who welcome the idea; so do the many competent to make subtle and sophisti- scientists impatient with what they see cated scienti½c judgments than a few as lawyers’ pointless wrangling over hours in a legal seminar would trans- well-known scienti½c facts. Indeed, form scientists into judges competent where mass torts involve vast numbers to make subtle and sophisticated legal of litigants on the same issue, where the judgments; and may risk giving judges science concerned is especially complex, the false impression that they are quali- and where hired scienti½c guns are en- ½ed to appraise specialized and complex trenched on both sides, court-appointed scienti½c evidence. experts may well be the best way to reach the right upshot (and more uni- As judges’ gatekeeping responsibilities form results than the kind of legal lot- have grown, so too has their willingness tery in which some plaintiffs win huge to call directly on the scienti½c commu- awards and others nothing)–especially nity for help. Since 1975, under fre 706, if judges learn from Judge Pointer’s ex- a court has had the power to “appoint perience about the pitfalls of choosing witnesses of its own selection.” Used in scientists to advise them, and about a number of asbestos cases between 1987 instructing those scientists on record- and 1990, the practice came to public keeping, conflict of interest, etc. attention in the late 1990s, when Judge Still, though the conclusion the Point- Sam Pointer, to whom several thousand er panel reached was almost certainly federal silicone breast implant cases had correct, it is troubling to think that just been consolidated, appointed a National four scientists–all of whom combined Science Panel to report on whether these this work with their regular jobs, and implants were implicated in the system- one of whom revealed poor judgment, ic connective-tissue diseases attributed to say the least, in signing a letter, while to them. In 1998, the four-member panel serving on the panel, to ask for ½nancial reported that the evidence did not war- support for another project from one rant claims that the implants caused of the defendant companies–were in these diseases. (Six months later, a thir- effect responsible for the disposition of teen-member committee of the Institute thousands of cases. More radically than of Medicine reached the same conclu- Frye’s oblique deference to the relevant sion.) The plan had been for the video- scienti½c community–more radically taped testimony of panel members to be even than Daubert’s (and Joiner’s and presented at trial; after the contents of Kumho’s) extension of judges’ gate- the report became known, however, and keeping powers–reliance on court- before the testimony had been tran- appointed scientists departs from the scribed, most of the cases were settled. adversarial culture of the common-law

Dædalus Fall 2003 61 Susan approach. Proponents have recognized only if Y caused harm to Z, given other Haack on this from the beginning: “[t]he expert desiderata of principle or policy: that it science should be regarded as an amicus curiae” is worse to convict the innocent than to (John Ordronaux); a court should have free the guilty; that constitutional rights the power to appoint “a board of experts must be observed; that legal resolutions or a single expert, not called by either should be prompt and ½nal; that people side” (Judge Learned Hand, 1901). So should not be discouraged from making have contemporary critics of the prac- repairs that, if made earlier, might have tice, such as Sheila Jasanoff, who com- prevented the events for which they are plain that it is elitist, undemocratic, a being sued; etc. We also like to think move in the direction of an inquisitorial that our adversarial system (under system. which a jury is asked to decide, on the Then there are the ripple effects of basis of evidence presented by compet- those disturbing dna exonerations, ing advocates, held to legally proper con- which have prompted not only re- duct by a judge, whether guilt or liability newed scrutiny of forensic laboratories, has been established to the required renewed concern about how lineups are degree of ) is as good a way as we conducted and photographs presented can ½nd to reach the desired balance. But to eyewitnesses, moves to videotape in- problems with scienti½c testimony terrogations, and so on–all, surely, wel- oblige us to think harder both about come developments–but also legisla- exactly what balance is most desirable tion to overcome obstacles to admitting and about the best means to achieve it. ‘new’ evidence, i.e., the results of new There is no question about the desir- dna tests on old material. Notwith- ability of prompt and ½nal legal deci- standing the law’s traditional empha- sions; think of totalitarian regimes sis on (in Justice Blackmun’s words) where people routinely languish in jail “quick, ½nal, and binding” solutions, without trial, or of Dickens’s Jarndyce v. some states have mandated post- Jarndyce. Nevertheless, if new scienti½c conviction dna testing, and others have work makes it possible to establish that extended or eliminated the statute of an innocent person has been convicted, limitations where dna evidence may be it seems obtuse to refuse to compromise available. ½nality in the service of truth. And, while it is salutary to remember that the “The basic purpose of a trial is the deter- brouhaha over recovered memories also mination of truth,” the Supreme Court prompted some modi½cations of stat- averred in a 1966 ruling. “Our system of utes of limitations, with dna analysis criminal justice is best described as a there really are the strongest grounds for search for the truth,” Attorney General such an adaptation of the culture of the Janet Reno af½rmed in her introduction law. to the 1996 National Institute of Justice There is no question, either, that trial report on dna evidence, Convicted by by jury is a vastly superior way of getting Juries, Exonerated by Science. So we like to at the truth than the trials by oath, or- think; but it would be more accurate to deal, or combat that gradually came to say that the law seeks resolutions that an end after 1215, when the Fourth Later- correspond as closely as possible to the an Council prohibited priests from par- ideal of convicting X if and only if X did ticipating in such theologically ground- it, or obliging Y to compensate Z if and ed tests. Our adversarial system is a dis-

62 Dædalus Fall 2003 tant and highly evolved descendant of Science in the ½rst English jury trials; but it is not the courts perfectly adapted for an environment in which key factual questions can be an- swered only with the help of scienti½c work beyond the comprehension of anyone not trained in the relevant disci- pline. We value trial by jury in part be- cause we think it desirable that citizens participate in public life not only by vot- ing, but also by jury service; still, though such participation is a desirable expres- sion of the democratic ethos, civics edu- cation for jurors hardly seems adequate justi½cation for tolerating avoidable, consequential factual errors. But we also value trial by jury for a more fundamental reason: the protec- tion it affords citizens against partial or irrational determinations of fact. Court- appointed experts are no panacea, and there are both legal and practical prob- lems to be worked out; but if, where complex scienti½c evidence is con- cerned, we can sometimes do a signi½- cantly better job of determining the truth with their help, adapting the cul- ture of the law in this way might afford better protection, and thus better serve the fundamental goal.

Dædalus Fall 2003 63