In Daubert, the Supreme Court correctly saw that lower courts had reviewed the admissibility of expert testimony and its foundation on the basis of a principle – the Frye “general acceptance” test – that had been superseded by the more liberal admissibility guidance of the Congressionally enacted Federal Rules of Evidence. At the same time, when it sought to articulate guidance for this activity, it heightened the gatekeeping duties of judges. In doing this, it entered intellectual territory that is not readily accessible to judges with their typical training. And, the Daubert case itself generates internal tensions. The trilogy of cases also creates tensions between them and poses problems for other courts interpreting them.

In what follows, I sketch some of these issues and consider some other problems the court generated by entering the intellectual terrain of epistemology, philosophy of science, and the nature of causal inferences. None of these is easy, but the Court explicitly or implicitly took them up anyway. The easy part of the Daubert opinion was the rejection of the Frye test. A much more difficult matter for the courts and for the rest of us is understanding the import of the decisions, key interpretations by lower courts as they wrestle with the Courts’ decisions and applications to science, and how they should guide the admissibility of expert testimony and its scientific foundation.

A tiny bit of history suggests that the Supreme Court's Daubert decision seems odd because it did not mention and seemingly disregarded its own precedent from a decade earlier in Barefoot v. Estelle. This decision had held that cross-examination and jury assessment of witnesses’ credibility and reliability were sufficient to protect a criminal defendant in a death penalty case against dubious and unreliable expert testimony that was widely criticized by the expert's own profession. By the time Daubert was decided in 1993, instead of merely rejecting Frye, as Chief Justice Rehnquist argued in dissent, and going beyond the plain language of the Federal Rules of Evidence, it created a “reliability” screen for expert testimony. This contrasted with Barefoot v. Estelle. And, of course, the Daubert review is more elaborate than a Frye review, but how extensive it is depends upon how the trilogy of cases is understood.