WILLIAM DAUBERT and JOYCE DAUBERT, Individually and As Guardians Ad Litem for
JASON DAUBERT, and ANITA DE YOUNG, Individually And As Guardian Ad Litem
For ERIC SCHULLER,
Petitioners,

v.

MERRELL DOW PHARMACEURICALS, INC., Respondent,

ON WRIT OF CERTIORARI TO THE UNITED STATES COURT OF APPEALS FOR THE NINTH CIRCUIT

BRIEF AMICI CURIAE OF
NICOLAAS BLOEMBERGEN, ERMINIO COSTA, DUDLEY HERSCHBACH, JEROME KARLE, ARTHUR LANGER, WASSILY LEONTIEF, RICHARD S. LINDZEN, WILLIAM N. LIPSCOMB, DONALD B. LURIA, JOHN B. LITTLE, A. ALAN MOGHISSI, BROOKE T. MOSSMAN, ROBERT NOLAN, ARNO A. PENZIAS, FREDERICK SEITZ, A. FREDERICK SPHILHAUS, DIMITRIOS TRICHOPOULOS AND RICHARD WILSON

IN SUPPORT OF RESPONDENTS

SUMMARY OF ARGUMENT

The courts below strove to ensure that opinion testimony that was the result of methods of inquiry not generally accepted as valid in a particular field of scientific endeavor would not become the basis of decision by a jury of individuals with no particular knowledge or experience in scientific or technical matters. We believe that this effort is of great importance and the goal salutary. Scientific terminology and scholarly credentials can, and often do, have an inordinate effect on a lay person’s perception of the authoritativeness of a scientist even when, or more precisely, just because, the lay person is usually not able to evaluate the intrinsic merit of the propositions advanced by the scientist. We do not suggest that scientists or other scholars belong to a caste superior to lay jurors, but we do believe that the scientific community as a whole is better able to evaluate the data and methods used by scientists.

The impact and influence of scientific rhetoric can easily sway and mislead a jury. Opinion testimony based on scientific techniques or methodology that are not generally accepted by the scientific community is not reliably valid. Testimony must be based on facts or data reasonably relied upon by experts in a particular field. Testimony not so grounded will not assist the trier of fact to understand critical facts in issue, and is more likely to mislead.

Publication and peer review, while not guaranteeing that particular methods of scientific investigation or data are accurate, are vital means of permitting the scientific community to see the data, to understand them, to test them, to evaluate the methods and show, either that the data or methods of a particular scientist are correct and appropriate and do lead to valid conclusions or, if such be the case, are not correct and thus do not lead to valid conclusions.

I.  PUBLICATION AND PEER REVIEW ARE SEEN ESSENTIAL FOR DETERMINING WHETHER SCIENTIFIC PRINCIPLES OR METHODOLOGIES ARE ACCEPTED IN THE SCIENTIFIC COMMUNITY

Publication and peer review are central to the scientific enterprise of corroborating or falsifying methods, data theories and conclusions advance by scientists to explain natural phenomena. We agree with Chubin, et al. that publication and peer review are not the final answer to the question of "What is truth?" While it is correct, as Chubin, et al. assert, that peer review journals do not replicate and verify the experiments, research and analytical techniques, or data reported in the papers submitted for publication, peer review does not end with publication – publication is only a starting point of peer review. The function of publication is to make data and methods available to the scientific community beyond the scientist’s inner circle and beyond the editorial staff and referees of the journal: publication exposes the work to the scientific world at large, which then has the opportunity to replicate and confirm, or fail to replicate and falsify, the scientist’s proposition. Of course, peer review journals do not "warrant" that the ideas or information contained in the articles they print are "accurate, valid, certain, reliable or true" . Indeed scientists do not assert that they know what is immutably true – they are committed to searching for new, temporary, theories to explain, as best they can, phenomena. It is nothing more than a red herring to argue that peer review does not guarantee "truth". Of course, the "mere fact" of publication does not mean that the ideas and information reported in an article are generally accepted by, or represent the consensus views of, the relevant academic community. Again, publication is the first step in obtaining a (temporary) consensus: Without publication other scientists would be unaware of the idea and so there would never be a consensus.

We disagree strongly with the assertion of Chubin, et al. that the fact that ideas and information have not been published in a peer review journal does not mean that they are not generally accepted. While failure to publish does not imply that the ideas are "generally rejected", or that they cannot represent "good science", it does mean that the scientific community has not had the opportunity to test and confirm the "discoverer’s" findings and ideas, or to satisfy itself that proper methods were used and that the methods and results represent "good science".

There is a major difference between the presentation of evidence for the sake of establishing "truth" in science on one hand and law on the other. Typically, in the legal process a specific conclusion is the goal of the evidence offered. Thus there is a strong motivation to be selective because of bias. Quite often, the experiments will have been conducted with a pending case or at least potential litigation in mind. On the other hand, scientific journals are typically concerned with progress in a particular field, and in making progress with sound methodology. The journal will let the conclusions fall where they may. Individual scientists who might have an interest in reaching a particular conclusion at the expense of rigor in methodology are challenged in the peer review process. One important advantage of publication and peer review, as compared with cross-examination and contradictory testimony of the adversary’s ex0pert, is that peer review by the scientific community at large enables scientists who are truly disinterested in the outcome of a particular litigation, the prospects of a particular product, the financial success of a particular industry, or even "justice", will have the opportunity to test, evaluate and either confirm or refute data, findings or theories. We submit that this neutral process is of great value both in science and in the courts.

Certainly, without publication and subsequent peer review ideas cannot become "generally accepted". Indeed, the notions of confidentiality or "privilege" and "general acceptance" are antithetical. The court of appeals below was correct in remarking that the work of "experts" prepared for the specific purpose of litigation, and not published or peer reviewed, cannot be said with any confidence to utilize generally accepted principles and methods.

Techniques can be considered widely accepted when they are used to produce results that are accepted time and again for publication in reputable reviewed media. The publication once of a new technique, no matter how esteemed the reputation of the author or the journal, gives it no inherent validity. That first publication simply says the editor and reviewers feel the article is of interest and does not immediately see why the methodology or theory will not work. The consequent exposure, then, of the technique to the scientific community allows for discussion and hopefully its replication. Its continued use, and acceptance in peer review journals, provides some assurance of validity, but does not provide any guarantee.

When experts are asked to draw a conclusion in a courtroom, their expertise must be established. Acceptability is established in science when a scientist produces results that have appeared in peer review journals and have stood the test of subsequent challenges. Publication is not a sufficient test by itself.

Scientific "orthodoxy" should be tested in the laboratory, in the meeting rooms and hallways at scientific meetings and on the pages of journals, and not in the courtroom. Ronald Bayer, et al. use the term "good science" . "Good science" is science that is done well, not necessarily science that produces the desired results. It is impossible to determine what results are correct. In that sense, good science is that science that undergoes the scrutiny of peer review, not only in the publication process, but after publication, when the scientific community examines the published data, findings and theory and then tries to replicate the data and the findings. Publication in a peer review medium is therefore a necessary but not sufficient condition.

Contrary to the conclusion of the Chublin amici, we submit that because of the technical, and increasingly technical, nature of much of the expert testimony offered in court, the jury and the cross-examination process are not adequate to ensure that scientific evidence has been derived from facts or data generally relied on by experts in the field. Publication and peer review do serve that important purpose.

If the courts below had used a phrase such as "universally recognized scientific truths", as Petitioners and certain amici at least imply, they would have adopted a clearly unacceptable criterion. But that s not what the lower courts in this case did. They simply recognized the central role publication and peer review play in the process of detecting and reducing the use of "bad science" in litigation.

It is common ground, we believe, that in science accepted "truth" is not a constant: that it evolves, either gradually or discontinuously. As Sir Karl Popper, the preeminent philosopher of science has asserted, the scientific enterprise starts with a deductive method to derive hypotheses which are then tested by observation or experiment. Karl R. Popper, THE LOGIC OF SCIENTIFIC DISCOVERY 32 (1959). For a theory to be "scientific", it must set forth an hypothesis that is capable of being proven false by observation or experiment and the data produced through this testing must be capable of replication. An hypothesis can be falsified or disproved, but cannot, ultimately, be proven true because knowledge is always incomplete. An hypothesis that is tested and not falsified is corroborated, but not proved. Thus, scientific statements or theories are never final and are always subject to revision or rejection. On this, we agree with Bayer, et al. Replicability, which is noted by Popper and others as the hallmark and guarantee of scientific acceptability, involves other scientists testing the accuracy of observations or of the predictions of an hypothesis. Scientific experiments are, of course, always expected to be replicable.

All of this is also relevant to law, since the basic principles of reasoning or logic are no different in the field of law than in science. However, the functions of law, and thus the propositions to be established by evidence and logic in the effort to construct a system of descriptive general theories based on particular data, law consists of a system of normative general rules that are individualized to apply to particular cases. This distinction engenders a number of significant differences.

An important difference between science and law is that the propositions to be tested in science are predictive while the facts to be proved in the legal process arise out of situations that occurred in the past and which cannot be repeated exactly. The legal process rarely has the luxury of being able to repeat experimentally a disputed chain of causation to corroborate the proffered hypothesis, even if, in some cases, it might theoretically be possible.

The legal process of adjudication – litigation – has highly structured rules and devices concerning the reception of evidence and the proof of propositions, which largely determine the proof available to the finders of fact in particular cases. The rules of evidence in federal courts are now embodied in positive law, the Federal Rules of Evidence (with corresponding provisions in most states), as well as in jurisprudential precedents. We agree with Petitioners that the basic principle of the Federal Rules of Evidence can be stated simply: all relevant evidence is admissible, except as otherwise provided, while irrelevant evidence is not admissible. But this apparent simplicity of structure, so heavily relied upon by Petitioners, is misleading. Nearly all the other rules of evidence are rules of exclusion on one ground or another".

The scientists on whose behalf this brief is filed believe that the Bayer brief, filed in support of Petitioners, although it may speak for one group of scientists, unfortunately embodies a fundamental misconception f the relationship between science and law, a misconception which permeates the brief and invalidates its conclusions. In its Summary of Argument, the Bayer brief states, inter alla: "Perpetuating the reign of a supposed scientific orthodoxy in this way, whether in a research laboratory or in a courtroom, is profoundly inimical to the search for truth." Throughout the Bayer brief there are similar expressions indication a failure to distinguish between the purposes of science in the laboratory and scientific testimony in court. Indeed, the principal point of the Bayer brief is stated to be that "Scientific inquiry, like the fact finding process in the law, is undermined by a categorical refusal even to consider views or analysis that challenge the supposed conventional wisdom."

Much of the argument of the Bayer brief consists of a lengthy dissertation on the history and philosophy of science a good part of which is correct in its original context, as when the Bayer brief states that "Galileo was persecuted by the inquisition for challenging the geocentric orthodoxy promoted by the Aristotelian scientists who dominated the academies and universities." However, this, and numerous other examples of scientific theories, perhaps once called heresies, which have now become accepted as true, with which Bayer and his colleagues regale us, are quite irrelevant to the issues confronting United States courts in the twentieth century. Galileo was prosecuted by ecclesiastical authorities in the seventeenth century, and was ordered not to write in support of the Copernican theory. The ecclesiastical authorities claimed to know "the truth" and asserted moral and divine authority to ban heresy. But United States courts do not have, and do not assert or claim to have, any authority whatsoever over the publication or promulgation of scientific observations, data or theories, and are not at all concerned with "heresy". A rule excluding expert testimony in litigation would not inhibit any new Galileo.

There are two distinctly different areas of activity and authority which are mistakenly treated as one by the Bayer brief. First, there is the process of scientific research and publication which concerns the formulation, corroboration, and advancement of scientific principles and theories. The corpus of this work constitutes the field called "science"; and progress in science depends entirely and exclusively upon the activities and opinions of scientists. Second, there is an entirely separate area, which is the field of legal process that concerns the adjudication of rights between particular parties. In the course of litigation issues arise involving the admissibility of certain types of evidence from many fields, including science; but the legal process does not establish nor attempt to establish scientific theories, principles or "truth" for the purposes of science, but only for the purposes of adjudication.

A critical difference is that while in science it is recognized that "truth" is extremely mutable, in adjudication "truth" for the limited purpose of resolving disputes must become final and immutable in a relatively short time. This concept of finality – essential to adjudication – is completely foreign to science.

The standards for the reception of evidence by courts in the course of adjudication are thus not the same as the standards that scientists use in accepting or rejecting new data or theories. Indeed, there are many fields of science with varying standards regarding both observations and theories; and purposes of inquiry, investigation, and calculation that have literally no evidentiary support.

On the other hand, in adjudication the law establishes its own standards of precision and reliability, and mandates that the "truth" be determined from admissible evidence.

With respect to scientific matters, a witness who is qualified as an expert in a relevant field may state his conclusions as to scientific knowledge, provided that his conclusions are based on facts or data directly observed by him or are of a type reasonably relied upon by other experts in the same field. That is basically what the rule adopted by the circuit court in this case requires. The rule applied by the courts below does not require that the conclusions reached by a scientific expert be generally acceptable, but it does require that the scientific principles and the methodology upon which conclusions are based must have gained acceptance in the relevant scientific field to be admissible.

It is a complete non sequitur, and false, to suggest, as the Bayer brief does, that the refusal of a court to receive testimony from a scientist has anything to do with the advancement or progress of science. As the Bayer brief correctly suggests, scientists are constantly formulating and publishing novel theories. Sometimes the theories are supported by substantial data and lead to testable substantial consequences, but often theories are supported only by insubstantial or questionable data or by no data, but only by calculation and speculation.

In deciding which expert testimony to admit and which to reject, the court must strike a balance between the risk of rejecting potentially relevant "good science" and admitting "bad science". The problem is that non-scientists, whether judges or jurors, often cannot distinguish between "good science" and "bad science". The expertise provided by the peer review process of experts in a field evaluating proposed theories and the procedure used to arrive at them is of great assistance in providing judges with a benchmark. By using publication and peer review as a standard, the determination whether particular principles and methodology have received acceptance within the scientific community is one which courts can make. The determination whether particular scientific conclusions are or are not correct or accepted by science is one which judges need not make.

There is superficial appeal to the argument in the Bayer brief that because "truth" in science often changes, the courts should not look to "generally accepted" principles as a standard for admissibility of expert testimony, but the Bayer amici are wrong in suggesting that research in the laboratory is the same as presentation of testimony in court. Bayer, et al. state some correct fundamental principles and then subtly extend them beyond their realm of applicability. Most important is the notion that in seeking the truth, one must thoroughly examine unorthodox ideas. We agree: in science one must never dismiss a conclusion that is arrived at through sound research techniques simply because it is in disagreement with conclusions that were previously generally accepted. The search for an explanation that reconciles such differences is Often the spur to scientific inquiry. However, the Bayer amici seem to be arguing on this basis that unorthodox results should there, per se, be acceptable in the courtroom. This is a stretch to absurdity.

It is how the conclusions are reached, not what the conclusions are, that makes them "good science" today. Conclusions, however divergent from conventional wisdom, that are arrived at by using sound scientific methods, should be considered both in the laboratory and in the courtroom. If, as the Bayer amici suggest, the courts have thrown out conclusions because they are not in conformity with conventional thinking, that would be wrong. If courts have thrown out conclusions because the techniques for achieving them were not generally accepted, as the lower courts in this case did, that is a correct course.

In any event, whatever determination a court may make as tot he admissibility of proffered scientific testimony in some lawsuit will have no impact at all on the progress or advancement of science. Thus the issue now before the Court does not involve questions of scientific stasis or progress. The issue in the present case is quite simply whether the law should impose liability on the basis of testimony that states a conclusion on a scientific subject that is based on methodology and principles that are not generally accepted by other scientists at the time of the testimony is proffered. A court can operate on no other principle, for if it admits conclusions based on methodology and principles that are not now accepted by the scientific community, but may in the future come to be accepted, it is operating in the realm of speculation. One cannot today say what may be accepted 50, 10 or even one year from now. Because litigation must terminate, but science intends to continue inquiring forever, the standards of discourse must be quite different.

As we read the decisions of the lower courts in this case and in other Bendectin cases, the federal courts have not excluded experts conclusions because they are not in conformity with conventional thinking, but generally have correctly refused to admit evidence proffered as scientific when it was based upon methods not generally accepted as scientific.

III.  PETITIONERS’ PROPOSED APPROACH IS INADEQUATE AND DANGEROUS

The interpretation of the rules of evidence as they relate to the admissibility of expert testimony in the fields of science advanced by Petitioners is completely inadequate, for it contains none of the safeguards of review, replication and evaluation by the scientific community that are essential parts of the process of acceptable scientific inquiry. It is not enough, we submit, to rely on cross-examination of scientific experts before a lay jury to separate the valid from the bogus. It is doubtful whether a lay jury would understand the intricacies or subtleties of sophisticated analysis and criticism. Moreover, the credentials of the expert are likely to cause the jury to credit his or her opinion regardless of the validity of the methodology used to arrive at the expert’s conclusion.

Petitioners, in essence, would require a court to allow testimony of any individual who possesses adequate formal credentials, regardless of whether the expert’s own investigation and analysis conform to scientific norms. It is to the scientific procedural safeguards that a court, this Court, must look to prevent unsupported conclusion – which may, nevertheless, be appealing to a lay jury – from being presented as, but are not really, the result of bona fide scientific investigation.

When testimony prepared for litigation is offered through an asserted expert, and the substance of the proffered testimony has never previously been published and tested in the scientific community, the party attempting to challenge such methods employed in preparation of such testimony have been widely criticized or rejected by the scientific community, or that the theories employed or the conclusions reached have been rejected or falsified by other scientists. The process of publication and peer review, while not "guaranteeing" the witness provide, at least, some significant assurance that other scientists in the field have had the opportunity to examine and test the expert’s methods, data and theories, to consider if they are right or wrong, and to refute them if they are wrong.

This is not an abdication or delegation of the proper role of the court or the jury to editors and staff of scientific journals. Rather, we submit, it is the very procedure that Fed.R.Evid. 703 contemplates when it requires that the facts or data upon which an expert bases his opinion, if not his own, be "of a type reasonably relied on by experts in the particular field in forming opinions or inferences". Petitioners are correct that Rule 703 "is designed to bring judicial practice into line with the practice of experts when not in court."" but are wrong in contending that the decisions of the lower courts are inconsistent with that rule.

CONCLUSION

Publication, peer review and replication are the practices of the scientists outside the courtroom as they proceed with their task of corroborating or falsifying scientific theories and conclusions. This is the method the lower courts insisted on as a prerequisite to admissibility of scientific evidence. The test used by the lower courts in this case is soundly based upon and is in conformity with the requirements of Fed.R.Evid. 703.

For the reasons discussed above, the judgment of the courts below should be affirmed.