Limiting the Subjective

-- By JieunPaik - 16 Feb 2012

Law as a Science

In Courts on Trial: Myth and Reality in American Justice, Jerome Frank writes about the subjective nature of the fact-finding process at the trial court stage. When a trial judge is presented with contradictory testimonies regarding the facts of a case, what kind of standard does she apply in order to ascertain which of the testimonies is accurate? Frank writes that very subjective, personal criteria, such as a judge’s values, biases, personal experiences, or the unconscious, guide how she determines the facts of a case to be so that what one judge determines to be the facts of the case can differ greatly from what another judge determine to be the facts. Because of such a situation, although many participants in the judicial system think that the practice of law is a science in which erudite knowledge of the “crisp and definite” legal rules will lead to sound prediction of how a judge will rule in future cases, such prediction cannot be had. This “law as a science” conception is founded upon the belief that facts, on which the legal rules are applied, are definite, impartial and lay out the stark contours of the reality which gave rise to the case. However, “facts are what the judge think they are.” Thus, the judicial decisions vary depending on how the judge understands the facts to be.

Frank criticizes the way in which participants in the legal process fool themselves into thinking that law is predictable and, thus, dependable. He urges one to see the practice of law as less a science and more as an art, by which I understood him to mean as a very human activity in which the personal natures of both counsels and judges shape the outcome of the case. He says that one should acknowledge the imbedded subjectivity of fact-finding and the corresponding unpredictable nature of the trial court decisions.

I agree with Frank in saying it is important to realize that the ability to predict judicial decisions is a myth.

It does not follow, on the argument you have given. Evidently, where testimony conflicts and judges or jurors must choose what witnesses to believe, there are inevitable uncertainties that would not exist if all testimony (or indeed all evidence of both kinds) on each material point in a lawsuit were uncontested and self-interpreting. (It is important to remember that circumstances as well as testimony must be interpreted in order to reach factual conclusions.)

But having said so much, one has not said anything about the overall predictability of outcomes, or the relevant qualities of predictability necessary to the practical as opposed to ideal management of risks, until one has both quantified the degree of uncertainty in fact-finding, compared it to other sources of relevant uncertainty, and qualitatively assessed its important to the kinds of legal prediction necessary to manage litigation risk. It might be that everything Frank says is true, and yet the phenomena he points at are comparatively unimportant.

It is really interesting to see how other aspects of modern society are viewed and understood to be imperfect and yet the legal system is still cloaked in mythic idealism. For example, people understand that politics, no matter how wide the smiles of the politicians, will be contentious and partisan. Yet it seems as if fewer people are willing to admit that similar attributes characterize the judicial system as well, fueled by the subjectivity of its participants.

But that's not what you just argued for, let alone what you proved. There's no similarity between the partisanship and rancor, or the insincerity, of electoral politics and the degree of variation in conclusions about whom to believe reached by similarly conscientious fact-finders in different cases involving witnesses offering similar testimony.

So what are we going to do about this situation? If the goal of the judicial system is to decide accurately who is right and who is wrong and to what degree, what are we going to do about the fact that two fact-finders can find the facts to be very different for the same case?

Not in the same case, but rather in different similar cases.

Could we at least make the fact-finding process be more objective?

That wouldn't help. Suppose you excluded testimony from your evidence altogether, and tried to find facts only the basis of circumstances without witnesses. You would still face a quantum of uncertainty in the nature of the fact-finder. The question would be the same as the question you have now: how to quantify that uncertainty and compare it to the other sources of uncertainty not to do with fact-finding, in order to determine the overall effect of that uncertainty on parties' ability to manage risks and the public's ability to trust the system's determinations.

How to Make Facts More Objective

To a degree, yes. Perhaps the utilization of more impartial methods of fact-gathering can limit the subjective element. However, I do not know how effective that would be. Just because one allows a mode of witnessing (such as videotapes) that is more impartial in its account than those of eyewitnesses

Who said videotape is more "impartial"? More importantly, who said videotaped testimony is better for fact-finding purposes, including determining whether someone is lying? What was the evidence for that proposition. I don't believe it on the basis of experience, so I would need to have it demonstrated in a fashion that overcomes my skepticism.

does not mean that it can effectively overcome the subjective element. There will still be subjective interpretation of that objective material.

There's nothing objective about video, any more than there is something objective in a still image. Everything about both has been manipulated. Everything about both is synthetic.

The value systems, biases, and the unconscious of the fact-finders are strong and pervasive; they can prevail against even the biggest rational obstacle in their path and lead their humans to what they believe to be the “right” answer.

What I propose is that the judicial system should implement procedural mechanisms that can minimize the impact of one person’s subjectivity. For example, what about the idea of allowing an appellate court to have a much less deferential standard of review for the facts found by the lower court, especially if done by a trial judge and not a jury?

Thus replacing the biases of those who actually dealt with the evidence live by the biases of those who had no contact with the testimonial evidence at all?

Such a mechanism can act as a safety net and allow factual errors to be more effectively rectified. What about the idea of allowing more than one judge to decide trial courts?

They're called jurors.

The implication behind mandating more than one judge for appellate courts and only one judge for trial courts seems to be to stress the importance of the accurate answering of legal questions; however, it seems to me that the ability to find the facts as accurately as possible is just as important as the ability to apply the correct legal rule in deciding which party gets a victory in the courtroom. Of course, in discussing these procedural possibilities I have not taken into account practical concerns, such as judicial efficiency. The goal is to generate ideas that could further and strengthen the process of accurate fact-finding. Assuming that having more than one perspective about a situation can allow for a fuller and more complete vision and understanding, these ways can hopefully limit the occurrences of critical factual errors.

You have an experiment before you, then: are facts more reliably found in jury trials than in bench trials? I very much doubt that professional observers would line up in that corner. Moreover, even bench trials involve collaborative determination of facts, because law clerks or legal secretaries review the record, and the parties submit their own proposed findings of fact, so that judges "sitting alone" are actually hearing argument on facts, and discussing those arguments with their own lawyers who were also present for the testimony.

This is a good first draft with some more thinking ahead of it. You've taken the easy parts of Frank's argument too easily, and dealt with the hard parts almost not at all. I've pointed to some of the questions editorial engagement should ask in the comments above. (1) What social science results over the past sixty years affect our approach to Frank's arguments? (2) How might the testimonial evaluation capacity of judges, that is, professional fact-finders, be improved and standardized by training? (3) What do we actually know about the variability in fact-finding based on assessment of witness truthfulness, and how does that compare to other sources of legal variability, including those elements under the control of parties? (4) What other principled goals we have for trials interact with our desire for controlling variability in the reception of testimony? In other words, if "truth serum" worked, would we administer it to witnesses?