Science and the Law
D. Allan Bromley
Science and the law were among the earliest and the most fundamental foundations on which our Nation was established. Thomas Jefferson was a lawyer and a self-trained scientist, while Benjamin Franklin was a scientist and a self-trained lawyer. Both were ranked among the leading scientists of their age and both clearly had an enormous impact on the development of law in the United States. Their science, of course, was highly applied and would today be called technologywhich both then and now drives the engine of the burgeoning economy that in the early days fostered the growth and stability of our young Nation.
We sometimes forget that in the colonial days in North America, there were parallel Spanish, French, and English colonies. Only the latter prospered (more or less) from the beginning. Historians credit this to the fact that the English colonies, from their very founding, had a particular penchant for resolving political controversies and achieving social order through legal means and legal structures. This was not the case in the Spanish and French colonies.
As the Nation grew and prospered, science and law went their separate ways, developing unique characteristics and institutions, but both solidly rooted in European traditions. Although there are many differences between the two, I will caricature only a few. For example, in my own area of physics and the physical sciences, we have the arrogance to believe that with the tools of mathematics and fewer than 20 natural laws, we can aspire to eventually understand the entire natural universe and its evolution. I have no idea how many laws the average lawyer deals with on a regular basis, but most certainly that number is vastly greater than 20. It is important, however, to recognize that scientists and lawyers mean quite different things when they use the word law. To a lawyer, law means rules established by properly constituted social authority; to a scientist, as I shall discuss below, a law is a statement (normally expressed in mathematical language) that is distilled from a great many observations of how nature functions and that is assumed to be both permanent and universal.
Mathematics is, in many ways, the universal and international language of science. Of particular importance for this discussion are the subfields of statistics and probability. This follows because a great many scientific measurements and observations are necessarily, to some degree, uncertain. Scientists generally use both statistics and probability theory and are particularly interested in a quantitative estimate of the probability that a particular experimental finding did not occur merely by chance. Mathematics also allows us to see the essential simplicity and, indeed, beauty in what would otherwise appear capricious or complex. For example, the shape of the arms of a spiral galaxy, the shape of Cape Cod, and the shape of the whirlpool that develops in your bathtub are all governed by the same natural law.
I recognize that all too many legal viscera tighten perceptibly at the mere mention of the words "physics" or "mathematics." This is unfortunate. But you should not forget that mathematicians and scientists return the reaction when faced with legal words.
In the early days, it might have been possible for science and the law to remain in splendid isolation, but in today's increasingly technological society, science, technology, and the law are inexorably drawn together. And the time has surely come when we need to better understand one another and how we work.
The Scientific Method
As I mentioned above, there are fundamental differences between law and science. In law, by definition (and I simplify almost to the point of caricature) there is always an answer, and at a time certain (such as at the end of the court action). In science, there may well never be an answer, and certainly no answer can be expected within any given time frame. And while both science and law seek to arrive at conclusions based on rational reasoning from evidence, the two disciplines define evidence in quite different ways. It is always somewhat jarring to scientists that lawyers consider evidence to be the words that a witness (perhaps a scientific expert) says under oath. A scientist would not consider mere words as constituting evidence, no matter how prestigious and respected the speaker. Instead, the scientist defines evidence as the observational data on which the expert draws a conclusion.
In (again) simplistic terms, the goal of science is truth and the goal of law is justice, but the approach to fact-finding in the two fields is very different indeed. In addressing a question, scientists make observations or measurements, and on that basis develop an hypothesis that explains what they observe. Unless they can use their mathematical toolkits to derive testable predictions from the hypothesis, however, it is worthless. To the extent that such predictions continue to be confirmed by new observations and measurements, the hypothesis is refined and eventually merits the title of "theory." The worst thing that can happen to a valid theory is that it eventually fails to reproduce some carefully established experimental fact, and must therefore be further refined. It is, however, not discarded. It is simply modified so that it can be considered as a special case of a more general, all-encompassing theory that explains not only the new results, but also all of the old ones. This, for example, is what happened when Newton's dynamics were recognized as a special case of Einstein's Theory of General Relativity. Scientists now recognize that Newtonian dynamics are entirely adequate for all situations involving low velocities, but when the velocity becomes comparable to that of light (as is the case when dealing with elementary particles and large, powerful accelerators) Newtonian dynamics can no longer explain what is observed. Einstein's General Theory of Relativity is required. A theory that cannot, in principle, be tested until one of its predictions is found to be false is, in fact, not even considered a part of science.
There are, of course, other ways of thinking that also attempt to get at the truth, e.g., religion, art, intuition, and even law. But scientific knowledge differs from ordinary day-to-day knowledge in that science seeks not only to explain how events take place, but also why they take place. Then it attempts to organize that knowledge systematically. And scientific knowledge is cumulative and progressive, and builds continuously on past understanding. As Isaac Newton pointed out, "the fact that I can see farther than others simply reflects the fact that I stand on the shoulders of giants."
Most people are aware of science only through its applicationswhat we would loosely refer to as the applications of technology. These have had an enormous impact on our entire society, on the quality of our lives, and indeed even on their duration. In 1955 the average life expectancy of Americans was 58 years; today it is 82 years. Much of that extension can be directly credited to new scientific understanding, and new technologies that allow us to respond constructively to this new understanding. Many of today's medical procedures would have been considered miraculous even a decade ago. And developments in electronics have brought about a revolution in computation and communication that has literally shrunk our world into a global village.
What I have described is what is generally referred to as the scientific method, but it bears emphasis that there is no single scientific method that adequately describes the work of all scientists. For example, some scientists do controlled experiments; others, such as astronomers, are generally unable to control the phenomena they study and rely instead on rigorous observations to test their ideas. For all scientists, however, experimental or observational information must be coupled with explanationwith answers to the question "Why?" As the Nobel Laureate François Jacob once wrote, science is an "endless dialogue between imagination and experiment."
Although science strives for orderly investigations, I would be far from candid were I not to admit that the scientific enterprise often does not proceed in any neat or tidy fashion. Many lines of investigation turn out to be blind alleys and many hypotheses must be abandoned. Perhaps even more important, we have often implemented technologies as soon as they became available, without adequate understanding of what some of the unwanted side effects might be (for example, our global environment has been the victim of many of these premature technological implementations).
Scientists and engineers do make errors, and sometimes they are entirely wrong. Sometimes even the best scientists fail to follow proper procedures. And some scientists, like all humans, are (to put it bluntly) sloppy at times. Nevertheless, if a scientific idea is considered to be an important one, future researchers will usually find any errors that may be in it and will rectify them. This is why the British philosopher John Ziman insists that "Knowledge that is not public knowledge, accessible to testing by anyone who might so wishin any way, and with any type of measurement or technology that might be availableis not, in fact, science, but rather priestly lore."
The Judicial Method
In contrast, a judicial inquiry is always bounded in time, because closure is needed. Moreover, there is a striking difference in the way the questions are selected. For example, the Article III Courts (the federal court system for my purposes) will hear only an "actual case or controversy," whereas in science any scientific question is open to research. The judicial inquiry stops when all the evidence has been presented. And the conclusion must be based on that evidence, even if from a scientific point of view, the evidence is far from complete or has been presented in a light most favorable to the proponent.
In an increasingly technological and litigious society, it is becoming ever more the case that the facts presented to the court involve large elements of science and technology. How are the judges and lawyers to decide on the validity and reliability of what is being claimed, and whether it should be admitted as evidence, given that very few of them have any formal training or experience in either science or technology?
Before even attempting to discuss possible responses to this question, let me describe a few examples of legal situations where the science was either ignored or misused.
The Misuse and Disuse of Science
A famous case (discussed by Ayala and Black) is the 1944 Charlie Chaplin paternity case where the court ruled that Chaplin was indeed the father of a girl whose blood type made his parentage totally impossible. Presumably this judgment was made on the ground that despite the scientific impossibility, the judge decided that the girl would have a better life with Chaplin as her father, than with someone with pockets not as deep. The definitive scientific evidence was simply ignored.
There has been much litigation regarding the alleged cancer-producing effects of low-frequency electromagnetic radiation from power lines (for example)so called EMF effects. It has repeatedly been concluded, however, by the National Academy of Sciences and other senior scientific bodies, on the basis of all available evidence, that not only is there no evidence for such effects, but also no known aspects of either physics or biology that could produce such effects. Perhaps the most compelling common-sense argument is that the use of electricity in the United States has increased by a factor of about 30 since 1950 and, over the same period, the incidence of all cancers (except those specifically linked to smoking) and most specifically leukemia in children (the cancer most frequently attributed to EMF) has decreased systematically. There is no evidence, to the best of any scientific judgment, that any power line or electrical apparatus has ever caused a human cancer.
Nevertheless, in 1993, New York State's highest court held that a claimant could seek damages for a drop in property value caused by public fear of a right-of-way for a high-voltage power line. The claimant was not required to prove any medically or scientifically reasonable grounds for the phobia concerning the effects of electromagnetic fields. The court felt that the economic question of the loss in market value could be resolved without being "magnified and escalated by a whole new battery of electromagnetic-power engineers, scientists, or medical experts." Phobia, here, is the operative word. The conclusion is simply that if phobia is sufficiently widespread, then the fact that it has been shown to be scientifically baseless is irrelevant. In the same year, by the time I left the White House, we as a Nation had already spent over $32 billion responding to this EMF phobia, and it is still alive and well.
Sometimes the fault is not with the court, but rather with the Congress that passed the law in the first place. The Delaney Clause of the Federal Food, Drug, and Cosmetic Act is a case in point. It requires that any substance that has been shown to be carcinogenic in animals must be removed from the marketplace. Pyrimidone is an organic chemical that has been shown to be mildly carcinogenic in animal tests, but since it had never been detected in foodstuff, there had been no problem. One morning in 1992, however, I received a frantic phone call from Bill Reilly, then the Administrator of the U.S. Environmental Protection Agency. Scientists in the chemistry department at Columbia University had just discovered how to detect pyrimidone at concentrations a thousand times lower than had ever been possible before. Reilly's problem was that every bottle of red wine contained pyrimidone at this low concentration. His question was, "What do I do now?" The law now required that all such wine be removed from the marketplacedespite the fact that pyrimidone had been present since prehistory in all red wine without causing any harm to humans. Fortunately, I was able to work with the Food and Drug Administration to get an effective waiver of the Delaney Clause in this particular case. But this is an example of laws written with the best of intention, but which fail to recognize that science and technology move on.
Sometimes, too, the court can be more impressed by the scientific messengerthe expert witnessthan by the scientific message. For example, in the case of Wells v. Ortho Pharmaceutical Corporation, it was alleged that an Ortho spermicide had resulted in severe birth defects in a baby girl. In awarding some $5 million to the mother and daughter, the judge stated openly that his ruling was based largely on his opinion of the opposing scientific experts rather than on any of the scientific evidence itself. With respect to one of the scientific experts testifying on behalf of the plaintiff, the judge said:
His opinion at trial was the same as the opinion that he previously had offered in his depositionHis detailed explanation of how he had ruled out other possible causes demonstrated that his opinion was the product of a careful, methodical reasoning process and not mere speculation. His demeanor as a witness was excellent; he answered all questions fairly and openly in a balanced manner, translating technical terms and findings in a common, understandable language, and he gave no hint of bias or prejudice.
In contrast, with respect to one of Ortho's major scientific witnesses, the judge said:
His criticisms of plaintiff's attorneys and of expert witnesses who testify for plaintiffs in malformation lawsuits, as well as the absolute terms in which he expressed his conclusions, severely damaged his credibility.
So much for the scientific evidence itself! The judge could have asked how extensive had been the testing of the spermicide in question; how many peer-reviewed investigations had found it to have teratogenic (birth-defect causing) effects and how many had found none. He could also have asked whether this spermicide was a close relative to any other chemicals proved to be teratogenic in humans. And the list goes on.
The Question of Scientific Evidence and Experts
Let me return to the fundamental question. How do judges and lawyers decide what is valid scientific and technological evidence and what is not? U.S. courts have been experimenting for much of the past century with guidelines for the role of expert witnesses in assisting the courts in such matters.
Various standard tests apply. For example, an expert witness might well be asked whether the science presented in evidence had been peer reviewed. That is, had it been published in a journal that made detailed reviews by several scientific peers who are recognized experts in the particular field in question before it was accepted for publication? Or had it been published only in commercial journals where the reviews are typically less rigorous, or in a trade journal where there may be no review at all and where the author may have been paid by its manufacturer to write an article or editorial supporting a particular product?
The idea that judges should bow to scientific experts was first articulated in 1923 in the famous Frye v. United States murder trial. The case was a simple one. At the trial, Frye's attorney attempted to present evidence through an expert witness that Frye had successfully passed a primitive lie detector test. But the opposing counsel objected and the court sustained the objection. On appeal this ruling was upheld and Frye was convicted.
The so-called Frye rule, that governed the treatment of scientific evidence in the United States for more than 70 years, comes from this far-from-transparent paragraph from the judge's brief opinion.
Just when a scientific principle or a discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone, the evidential force of the principle must be recognized and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.
This general acceptance criterion has established itself as an attractive shorthand, but it is deceptive. Unfortunately, in today's courts, for sufficient money, one can usually find at least a few scientific experts who will testify to almost anything. All too frequently we find the consensus view of the scientific community being presented on the one hand and the views of experts who represent either themselves or, at best, a very small minority of the scientific community on the other. All too often the media report such situations as ones where "the scientific community is divided," conveying the impression that it is equally divided!
New U.S. Supreme Court Opinions
In the past 5 years, the U.S. Supreme Court has issued two important opinions addressing the role of federal trial judges and ruling on the admissibility of expert testimony. The Court's new guidelines for trial judges were written in response to cases that clearly involved scientific evidence. In the first, the issue was whether a particular drugBendectinused by pregnant women caused birth defects in their babies. In the second, the issue was whether polychlorinated biphenols (PCBs) in electrical transformers had caused an electrical worker to contract lung cancer.
In the first case (Daubert v. Merrell Dow Pharmaceuticals, Inc.) the Court stressed that a trial judge must determine "whether the reasoning or methodology underlying the testimony is scientifically valid." Evidentiary reliability, held the Court, depends on scientific validity. To guide trial judges in their new gatekeeping task, the Court briefly discussed four nonexclusive factors: 1) whether the experts' hypothesis is falsifiable, and whether it has been tested; 2) whether the theory or technique has been subject to peer review; 3) the known or potential rate of error; and 4) whether the experts' methodology and reasoning is generally accepted in the relevant field. The Court noted that this inquiry should address the methodology, not the conclusions, of an expert.
In the second case (General Electric Company v. Joiner) the Court went further and noted that conclusions cannot be absolutely separated from methods; the two must be considered together. That is, if an expert's conclusion is at odds with generally accepted knowledge in the expert's field, the conclusion should be scrutinized very closely. The Court also emphasized that an expert must explain both how and why the conclusions were reached. More than the expert's say-so is required.
AAAS joined with the National Academy of Sciences to file an amici brief in the Daubert case. They both urged the Court to require trial courts to apply "the same trial criteria that scientists themselves regularly apply in picking and choosing the theories, explanations, and methods upon which they build their own work." In many ways, the Supreme Court decisions in Daubert and Joiner reflect precisely this approach. Simply put, the Court has required that scientific testimony must, in fact, be scientific.
As a consequence, federal trial judges (and many state judges as well) must now articulate their admissibility determinations in a rational manner that reflects a more thorough understanding of scientific issues and reasoning. Moreover, attorneys must now develop a better understanding of the scientific issues in their cases, and they must frame their arguments to address the new guidelines under which the trial judges are now operating.
Possibilities for Future Improvements
There have been many suggestions as to how better to bring science to the service of the courts, and how better and more fairly to use expert scientific witnesses. All have had rather obvious shortcomings, however.
Among the most radical of these suggestions is one that would draw on the practices of the European Civil Law jurisdictions. It requires that the expert scientific witnesses be appointed by the court and be answerable to the judge alone, rather than having them appear on behalf of one of the litigants. It bears noting, however, that under Rule 706 of the Federal Rules of Evidence, U.S. federal judges already have the power to appoint impartial expert witnesses, but they rarely used it, perhaps in deference to tradition and perhaps to avoid usurping a jury's fact-finding prerogatives.
In February 1998, however, Associate Justice Stephen Breyer of the U.S. Supreme Court, endorsed a Demonstration Report proposed by AAAS to be implemented over a 5-year period. Assuming that foundation support for this project will be forthcoming, it will provide a slate of candidates to serve as court-appointed experts in cases in which the court has determined that the traditional means of clarifying issues under the adversarial system are unlikely to yield the information necessary for a reasoned and principled resolution of the disputed issues. AAAS intends to seek nominations of qualified experts from the national scientific and engineering societies. Experience gained through the Demonstration Project will serve as the basis for recommendations regarding the appropriateness and scope of a permanent institutional mechanism that will link the federal courts, and ultimately the state courts, with the scientific and engineering communities.
Ultimately, the question at issue is one of translation between the ingrown habits, traditions, and cultures of the law and of science. And translation is a complex and sometimes tricky business. Let me conclude with an anecdote told to me by Dr. Guillermo Soberon, the Science Advisor to former President Salinas of Mexico.
It seems that a Mexican bandit robbed a small southern Texas bank and escaped back to Mexico with the loot. Unfortunately for him, a Texas ranger was in hot pursuit and caught up with the bandit in a small village south of the border. But there was a problem. The ranger could speak only English and the bandit only Spanish. So, in time honored fashion, the ranger found the most distinguished looking elderly resident of the village who could speak both languages and deputized him to translate. The following conversation ensued:
Ask him what his name is.
At this point, the ranger pulled his pistol, put it to the bandit's head, and said, "Try again." This time the answer came back in very soft Spanish, "I hid it in the village well." And the translation came back with equal speed, "José says he is not afraid to die!"
The point of this little anecdote is simply that you would be wise to ensure that your translator shares your agendasomething that all of you would be wise to remember in dealing with experts in the law.
D. Allan Bromley is Sterling professor of the sciences and dean of engineering at Yale University. This article is based on remarks he made August 2, 1998 during the 1998 Annual Meeting of the American Bar Association in Toronto, Ontario, Canada. The author wishes to acknowledge the assistance and education received from Ms. Deborah Runkle and Dr. Victoria Sutton during the preparation of his remarks.