At a recent conference of California state judges, I was on a panel discussing the use of scientific evidence. Climbing up on my soapbox, I declared, "At the end of the 20th century, it is an embarrassment that most judges cannot even distinguish a mean from a median from a mode." Charles Nesson, the Harvard Law professor who moderated the session, looked out to the audience of about 200 judges and asked them to raise their hands if they could define those terms. About 15 hands went up. Another 10 judges raised their hands tentatively, as if hoping not to be called on.

I understand that most Americans cannot define these statistical terms. Surveys indicate that most Americans do not know who James Madison was, either. But judges are not most Americans. Just as we expect judges to be able to identify Madison, whose critical contributions to the drafting of the U.S. Constitution are taught in every law school, we should expect them to have a basic understanding of science. (For those of you who don't remember junior high school math, the mean is the arithmetic average of a group of numbers, the median is the middle score and the mode is the most frequently occurring number.)

In fact, scientific issues are litigated far more often than constitutional ones. Judges, the last great professional generalists, hear an incredible array of cases that hinge on scientific issues, such as the reintroduction of the gray wolf in Yellowstone, the authenticity of a claimed Rembrandt and the possible carcinogenic effects of electromagnetic fields. A recent study found that more than 80 percent of civil cases involve at least one expert witness. Most of these witnesses base their expertise ostensibly on science.

Complicating the judge's task, "science" often comes into the courtroom distorted by the adversarial process which, at best, spins it to the prosecution's or the defense's advantage, and, at worst, perverts it beyond recognition. In some cases, there is no science to speak of--the expert's view is based merely on his experience.

Yet judges--and many readily admit this--are sorely lacking in the skills necessary to closely analyze the bases for scientific expert testimony. (For many lawyers, this disability is established early: Many of my law school students have told me that a principal reason for choosing the law was a desire to avoid math and science.)

In an effort to remedy this gap, judges have begun taking classes on scientific evidence. I teach some of these classes, and I hear judges saying that they don't have the background, the time or the inclination to learn the minutiae of science. They are frustrated by the enormity of the task. In addition to these sorts of classes, other remedies are coming to the fore, including scientific reference manuals designed for judges. Especially promising is the Court Appointed Scientific Experts project of the American Association for the Advancement of Science (AAAS). Beginning this fall, the project will recommend impartial scientists whom judges can call upon as experts for the court. They will be available to help judges understand the science well enough to make decisions regarding the admissibility of the evidence and, in some cases, will testify before the jury.

The AAAS proposal comes none too soon.

In 1993, the Supreme Court ruled that judges are obligated to scrutinize scientific evidence. In Daubert v. Merrell Dow Pharmaceuticals Inc., the court held that the Federal Rules of Evidence require judges to be "gatekeepers" who "must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable." Most states have followed the federal lead. Judges, therefore, can no longer passively permit experts to testify merely on the basis of their professional degrees or because one side or the other was willing to employ them.

Impartial court-appointed experts can help judges immeasurably in fulfilling this new role. As tutors sitting by the judge's side, they will give judges greater confidence in dealing with unfamiliar scientific subjects. And since, under the AAAS plan, experts will be appointed by the bench, they will be largely removed from the adversarial process, allowing them to be impartial advocates for truth and fairness rather than for one side or the other. For me, this is one of the beauties of the proposal.

An expert retained by one litigation team or the other very quickly becomes a strategist for that side. It happened to me the one time I served as an expert witness. It all started with a phone call from a former student who remembered my lament in evidence class that, based on the available research, handwriting identification is little better than tea-leaf reading. She was assisting on a criminal trial in which the government intended to introduce expert testimony on handwriting identification. The defendant was accused of drug possession with intent to distribute, and the evidence included incriminating notes, which the government expert alleged were written by the defendant.

My only task, I thought, was to testify at a preliminary hearing before the judge regarding whether the jury should hear the government's handwriting expert. I had every intention of remaining neutral on the case itself and hoped to be seen merely as the font of "truth" on handwriting analysis expertise. It didn't take long, however, for the proceeding to become adversarial. The presiding federal judge denied our motion to exclude the government's expert. He did, however, express dismay at the lack of research supporting the claims of handwriting experts and agreed that the expert would have to testify as a specialist based on his experience, rather than as a scientist relying on research. Changing the expert's label, he ruled, would make the jury less likely to be swayed by the expert's claim to be part of the scientific tradition.

The defense counsel then proposed that I testify to the jury on the lack of research supporting handwriting experts. The judge seemed skeptical, but in time accepted that my graduate training in statistics and research methods qualified me to discuss these issues. And so I became an anti-expert expert. My expectation that I could remain neutral disappeared. I had joined the defense team. I testified that little research had been done on handwriting and that experts today continued to use techniques that were developed nearly 100 years ago. Experience with a subject--the claim upon which handwriting expertise mainly rests--is not proof of expertise. If it were, the Earth would still be considered flat.

Then the cross-examination began. The assistant U.S. attorney wanted to know: Wouldn't the expert's experience with handwriting assist the jury to determine who authored the note? Shouldn't experience be worth something? Common sense suggested that it should. I thought so too, but as the questioning continued I became less and less inclined to acknowledge the prosecution's common sense: I was on "the other side."

I have spoken to many scientists who have served as expert witnesses and their courtroom experiences are similar. Skepticism, questioning the strength of your own hypotheses, keeping an open mind about alternative explanations and even common sense are all part of the scientific enterprise. In court, they become weaknesses exploited by the opposition.

As I suffered through cross-examination, I wanted to call a halt to the proceedings and shout out, "This is not the way science is done!" Since judges generally frown on outbursts in their courtrooms, I remained silent.

Impartial experts of the kind envisioned by the AAAS proposal have already shown their effectiveness in complex civil cases such as silicone breast implant litigation. In 1992, Pamela Johnson won a verdict for $25 million because a Texas jury found, based in part on expert testimony, that implants caused her autoimmune disorder. Later, two federal courts in other cases employed neutral expert panels and reached the opposite conclusion: Both determined that research indicates little or no relationship between silicone implants and autoimmune disorders. While these panels were created ad hoc by the judges in those cases, they operated much like those proposed by the AAAS. Such experts obviously do not guarantee that judges will always get it right, but they should do much to improve the odds that they will.

The presence of court-appointed experts will have other salutary effects. Most immediately, it seems to me, it will help juries and judges understand the issues more quickly, likely leading to faster settlements and lowering litigation costs. I also believe that court-appointed experts could keep the opposing sides' experts honest: They are less likely to make heroic claims when an authority in their field is sitting next to the judge. Most importantly, though, they will increase the science quotient of every judge they work with. And that is a benefit with lasting effect.

David Faigman is a professor at the University of California Hastings College of the Law in San Francisco. His book "Legal Alchemy: The Use and Misuse of Science in the Law" will be published in October by W.H. Freeman & Co.