Stephen Palley
Sep 6, 2018 · 7 min read

How do you prove something did or didn’t happen? If you’re in court, you use evidence.

Evidence can include physical evidence (“the murder weapon”, “the contract”) and witness testimony (“Q: What color car was the defendant driving? A: Blue.”)[1]

Evidence must be admissible to be used at trial. Judges decide what is or isn’t admissible. Cases and legal code guide them. In American federal courts, The Federal Rules of Evidence (“F.R.E.”) apply.[2]

At the end of a trial, if the judge agrees that enough evidence has been presented, s/he will (1) instruct the jury on the law and (2) send it off to deliberate with jury instructions. Those instructions will describe the elements of the parties’ claims and defenses, as well as other legal principles that a jury can consider while deliberating.[3]

Evidence rules serve procedural functions (“this is admissible/this isn’t”) and substantive functions (“this is fair/this isn’t”). The procedural functions derive their purpose, in part, from substantive rights that they protect. The search for truth in a particular case is supposed to be balanced with a commitment to fairness in all cases. Not all evidence gets to a jury.

There are two general categories of evidence. Direct evidence is evidence that if believed proves a fact, without requiring any inference. Circumstantial evidence is evidence that can also be used to prove a fact, but it requires inference. An example may help here. Assume you want to prove it was raining on a particular date and time, at a particular place. If a witness testifies that they they were there at the relevant date and time and got rained on, that’s direct evidence it was raining. If they testify that they were in a building at the relevant location at the relevant date and time, didn’t see the rain, but saw a bunch of wet umbrellas and raincoats, that is circumstantial evidence that it was raining. An official document containing a contemporaneous record of rain might also be admissible, assuming the proper foundation can be established.

Direct evidence isn’t better or worse than circumstantial evidence.[3] While this might seem odd, it makes sense when you think about it. The fact that a witness says they something happen doesn’t make it so. People lie. They can also can be wrong without being liars. Their credibility can be challenged on cross-examination.[4] Generally speaking, it’s up to a jury to decide how much weight to give either kind of evidence.

The rules also limit what kind of evidence can be used. For example, the hearsay evidence rule says that an out of court statement can’t be used to prove its own truth (“the truth of the matter asserted”). The idea here is simple: if you’re going to try to prove something using a witness’ words, the actual witness should be present in court so that the person against whom those words are being used is able to question them. In the United States this right is embodied in the 6th Amendment’s confrontation clause.

There are many hearsay exceptions. They reflect legislative or court made judgments about the reliability of certain out-of-court statements. A “dying declaration” is a classic hearsay exception, premised on the notion (sensible or not) that someone who is about to die is unlikely to lie. Under this exception, a witness could testify about an out-of-court statement by a dying person, and the statement could be used for proof of the assertion (After establishing X was about to die, “Q: What did W say? A: He said he saw the defendant kill V.” You could of course cross-examine the witness who is offering the hearsay evidence and question their credibility. But a hearsay objection would be overruled).

You can also use out of court statements for non-hearsay purposes, to prove something other than the assertion in the statement. For example, a witness might testify that someone said it was raining, and the testimony could be used to explain why the witness put on a rain-coat, not to prove that it was actually raining (“Objection: Hearsay. Response: Your honor, goes to subsequent conduct. Ruling: Over-ruled”).

As my first year evidence professor explained on the first day of class, all admissibility questions come down to (legal) relevance. Most other evidence rules are built upon it (hearsay, among them). Relevant evidence makes a fact you need to prove or defend your case (a “consequential” fact) more or less likely to be true. (See F.R.E. 401). It’s also called “probative” evidence. Although the rules don’t say this, exactly, hearsay evidence that isn’t subject to an exception isn’t legally relevant, so it’s not admissible.

Evidence can be “probative” and at the same time cause undue prejudice. If a piece of evidence’s probative value is outweighed by the risk of undue prejudice, it isn’t admissible. All evidence that helps one side and hurts the other is prejudicial, of course. The key is unfair prejudice. The balance between the two appears in Federal Rule of Evidence 403, which guides a trial court’s ruling: “The court may exclude relevant evidence if its probative value is substantially outweighed by a danger of one or more of the following: unfair prejudice, confusing the issues, misleading the jury, undue delay, wasting time, or needlessly presenting cumulative evidence.”

Some evidence is so prejudicial that it has its own separate rules. Character evidence is an example. Proving or disproving a fact based on evidence of a witness’s character is a risky proposition, and character evidence is carefully circumscribed as a result (though sometimes admissible, depending).

A case in point is F.R.E. 404(b) which provides that “[e]vidence of a crime, wrong or other act is not admissible to prove a person’s character in order to show that on a particular occasion that person acted in accordance with the character.” In other words, you can’t use the fact that the defendant robbed banks in the past to prove that they did it again, because it’s in their nature. There are some important exceptions, as always. (“motive, opportunity, intent, preparation . . . [etc.]” See F.R.E. 404(b)(2)). That said, each case is supposed to stand alone. Prior bad act evidence can have such a prejudicial effect on a jury (“they did it before, they must have done it again”) that it can get it the way of the actual truth. (Again, I am glossing over some important nuance here, but this isn’t an evidence treatise.)

The rules of evidence are an important part of the American dispute resolution system, one which is built of truth-finding through advocacy. The rules are elegant and flexible, but they are by no means perfect. Sometimes they lead to incredibly unjust results. One flaw is external to the rules — their application sometimes depends on the quality of the trial lawyers who are using them. In an adversarial system, the quality of your lawyer can have a significant outcome on your case. Formal licensing requirements establish baseline competency requirements for lawyers but (with some exceptions) there are few U.S. courts that require special admission in order to try a case. Outcomes can also depend on human judges and human juries. There’s a problem with a truth-finding system in which the same facts and the same law can result in different decisions.

I can hear some of my programmer friends saying “See! This is exactly why we need software to come in and fix things. Law is too complicated, too messy, and leads to too much injustice. Code is clean, unambiguous, testable and predictable.” It is easy to leap from the fact (1) humans are the flaw in the system to an assumption that the answer lies in non-human (2) software that knows all of our rules and will applies them uniformly. The leap takes one over a yawning precipice, full of peril, and a potential risk of even greater unfairness than our current system causes.

Algorithimic justice runs the risk of taking systemic unfairness and placing it into a black box. There’s no guarantee that human created “robot” judges will be any better at dispensing justice than their human creators. They might even do worse. While our rules of evidence and procedure are far from perfect, they are also far better than any other generic system of proof I’ve come across. It is a one-size does not fit all framework by design. Fix, yes. Replace, no. I’ll allow that this is editorializing ipse dixit, but the way to even the playing field further is make better judges, better lawyers, and better tools, building on frameworks that we already have.

One area where trial lawyers can use help is in authentication and admission of evidence. This is low-hanging fruit. From a trial lawyer’s standpoint, blockchain technology might make proving that things did or did not happen easier, and less dependent on a trial lawyer’s mastery of technical admissibility requirements. It shouldn’t be necessary to create new laws to allow that to happen thought — the rules of evidence are already more than sufficient.

[1]. Sometimes witness testimony is necessary for physical evidence to be “admitted”.

[2]. Many of these rules have common law antecedents that predate the F.R.E. A large body of caselaw interprets the rules. Some American state courts have adapted versions of the F.R.E. Others have their own evidence codes. Many of the same principles apply, in state or federal court.

[3]. Here’s a simple example: “If you find by a preponderance of the evidence that x, y and z are true, then you should enter a verdict on Count I for the Defendant.” (“A seprate instruction would then define “preponderance of the evidence.”) Once instructed, the jury then decides whether they have been presented with enough evidence to conclude that x, y or x are or are not true. Non-jury bench trials are different. For purposes of this post, I am going to limit the discussion to jury trials.

[4]. F. Lee Bailey’s cross-examination technique from the Simpson criminal trial is a case in point: https://youtu.be/gVoIz2zNX9U?list=RD_vIyfzbPCqY.

Stephen Palley

Written by

Itinerant slant rhymer. Lawyer. “I contain multitudes”.