Scientific Misconceptions Among Daubert Gatekeepers: The Need for Reform of Expert Review Procedures

Abstract

JAN BEYEA [*] DANIEL BERGER [**] I INTRODUCTION The Supreme Court's opinions in Daubert v. Merrell Dow Pharmaceuticals, Inc. [1], General Electric Company v. Joiner [2], and Kumho Tire v. Carmichael [3] contain two inconsistent views of science. In some places, the Court views science as an imperfect "process" for refining theories, whereas in other places, the Court views science as universal knowledge derived through "formal logic." The latter view, long out of favor with philosophers and historians of science, comports with the current cultural vision of science and is likely to be adopted by district and appeals court judges, without vigorous "education," or until such time as higher courts recognize that the two views need to be synthesized into a consistent whole. The interpretation of science as seamless logic puts an undue burden on scientific experts who are challenged under Daubert. In fact, Daubert, as interpreted by "logician" judges, can amount to a super-Frye [4] test requiring universal acceptance of the reasoning in an expert's testimony. It also can, in effect, raise the burden of proof in science-dominated cases from the acceptable "more likely than not" standard to the nearly impossible burden of "beyond a reasonable doubt." Instead of relying solely on the second view, courts should combine the two. A synthesis of the two views of science can be achieved by recognizing that subjective assumptions and inferences can never be completely eliminated from expert testimony. As a result, expert testimony always amounts, in effect, to conditional statements. An expert's statements can be considered "reasonable"--or likely, or beyond a reasonable doubt--if, and only if, the assumptions and inferences made by the expert are considered reasonable, or likely, or beyond a reasonable doubt. This "Bayesian" approach [5] may offer a pragmatic synthesis of the tension in the Daubert decisions for science-dominated cases, provided that courts can find ways to judge fairly the reasonableness of underlying scientific assumptions. This article offers a number of proposals to meet this goal, along with other suggestions for reform of the Daubert review process. The gatekeeping function assigned to judges by the Daubert decision and strengthened by the subsequent Supreme Court cases Joiner and Kumho Tire, has been both vilified and praised in the legal literature. [6] This article supports the critics of Daubert, who see science as a contentious process, rather than a catalog of truths, and who argue that courts are now demanding more of individual scientists and engineers than is expected of them in their own research and practice. Moreover this article follows cues about science from world-renowned scientists, including Neils Bohr [7] and Norbert Wiener, [8] and influential philosophers of science, such as Thomas Kuhn, [9] all of whom have written about the limitations of individual scientists and the subjective elements inherent in the scientific process. The competing school, who might be considered the intellectual descendants of John von Neumann, the inventor of the basic methodology used in the digital computer, would likely disagree. [10] Von Neumann saw sci ence as a set of universal facts deduced by logic, thereby placing scientific knowledge within a formal logical structure that he viewed as primary and universal. [11] This article proposes that the court interpretation of Daubert as requiring "seamless logic" fails to recognize that science is an imperfect process frequently built on assumptions and inferences that simply cannot be proved by the formal logic approach Daubert seems to require. Part I outlines the two competing schools of the philosophy of science and examines the roles these two schools have played in litigation. Part II distinguishes Daubert, which tends to follow the "formal logic" approach, from the "process" view of science, showing both the flaws in Daubert's approach and the ways in which a shift to the "process" school would render Daubert more acceptable. …