It is not hard to guess why courts might see fit to institute bans on loaded words at criminal proceedings. The Constitution protects a criminal defendant’s right to a trial at which jurors must presume that the defendant is innocent unless and until persuaded of her guilt beyond any reasonable doubt. As DNA testing has established, juries sometimes convict defendants of crimes that they did not commit. The obvious inference is that something has gone wrong in the courtroom and that this “something” explains why juries are concluding that innocent people are guilty of crimes. Word bans reflect the conclusion that that “something” is the failure of criminal juries to apply the standard of proof as rigorously as they should and that this failure may result from the style of witnesses’ narratives, including the particular words they use.

The notion of regulating the manner in which the parties may present evidence to the jury is hardly revolutionary. Prior to the enactment of the Federal Rules of Evidence in 1975, in fact, federal law contained a word ban of sorts, in the form of an “opinion rule.” The opinion rule required that witnesses recite only facts, but not opinions, to the jury. Used in this context, an “opinion” is not a point of view about a disputable proposition (such as “vegan strawberry milkshakes are delicious”). It is instead a conclusion (such as “the cow was dead”), as opposed to a recitation of the facts on the basis of which one has drawn the conclusion (such as “the cow’s heart was not beating,” “her skin was cold and her eyes glazed,” and “maggots had begun to colonize her flesh”).

Witnesses ordinarily speak in terms of “opinions,” rather than “facts.” It was therefore frustrating for witnesses to be stopped mid-testimony and told to “give us facts rather than opinions,” as the common law approach required. The result would often be to confuse a witness and thereby make it more difficult for him to tell his story.

In addition, it is not clear where the line between facts and opinions lies. After all, one could characterize any factual claim as a conclusion based on observation. This is what we mean when we say that our eyes play tricks on us – we automatically strive to make sense out of what we see, hear, smell, taste, and touch, and the sense we make involves the drawing of an inference or opinion. When you say that you saw John Doe, for example, what you really mean is that you saw a person with many of the same features shared by Doe and that you drew the conclusion that it was Doe. But that is not how anyone usefully conveys recognition of a person’s face.

Realizing these flaws in the opinion rule, Congress passed Rule 701 of the Federal Rules of Evidence, which says the following:

“If the witness is not testifying as an expert, the witness' testimony in the form of opinions or inferences is limited to those opinions or inferences which are (a) rationally based on the perception of the witness, and (b) helpful to a clear understanding of the witness' testimony or the determination of a fact in issue, and (c) not based on scientific, technical, or other specialized knowledge within the scope of Rule 702.”

In other words, as long as opinion testimony by a lay witness (whose testimony is more tightly controlled than that of an expert) is rationally based on her perception and helpful to the jury, it is admissible. Rule 701 permits witnesses to talk in a more natural fashion, rather than being interrupted with requests that would likely make little sense to a non-legally-trained person attempting to convey a factual narrative.

In an attempt to protect defendants’ right to the presumption of innocence, some courts have effectively enacted the limitations of the “opinion rule” in the form of a loaded-word ban. Where a police officer might want to testify about a crime victim’s statement, the rule requires that he use words other than “victim” instead. Only the jurors, on this reasoning, get to decide whether a particular complaining party is indeed a crime victim. Until they have so concluded, the use of the word “victim” is prejudicial and inappropriate.

On this logic, a witness testifying that the defendant “raped” her is similarly out of line. Only the jury can decide whether the complaining witness was in fact raped. Until it does, the witness must use more neutral language, such as “the defendant had intercourse” with her, while she resisted and told him loudly to stop.

This logic, however, is flawed, for the same reason that the earlier opinion rule was flawed. Witnesses – especially lay witnesses – speak in terms of factual conclusions, rather than basic building-block facts. To ask them to do otherwise is as likely to confuse them and produce incoherence as it is to enlighten the jury. Furthermore, any accurate description of a crime by a prosecution witness will necessarily draw conclusions about what occurred, and these conclusions will likely favor the prosecution over the defense, no matter how neutrally-phrased.

People understand what a witness means when she says the defendant raped her, in other words, and a more “neutral” rendition of the same facts will still convey to the jury a prosecution version of events: i.e., that the defendant forced the complainant to have intercourse. Jury members will know, moreover, , especially when instructed at length by the judge on the law, that they need not take the witness’s word as to what happened; that is the very question they are charged with deciding, based in part on their assessment of the witness’s credibility.

One fundamental problem in the “word ban” approach to protecting the presumption of innocence is its implicit assumption that witnesses are required, in their testimony, to presume that the defendant is innocent. This assumption is false. The victim-witness in a rape case is not required to “presume” anything about what occurred. She was present at the scene and knows what happened.

It is the jurors who must presume the defendant’s innocence, and this means only that they must find the defendant “not guilty” as long as they have not heard sufficient evidence to persuade them of his guilt beyond a reasonable doubt. Part of the evidence that may or may not be sufficient will consist of witnesses, such as the victim herself, who describe their own observations as clearly as they can. If the jury is persuaded that a witness has spoken honestly and accurately about what she accurately perceived and remembered, then it will, properly, consider such testimony as proof of the defendant’s guilt.

Some critics of the word bans have invoked the First Amendment freedom of speech, arguing that witnesses have the right to describe what occurred without government censorship in the form of word bans. Such critics, however, are mistaken. Witnesses do not have First Amendment rights to free speech in the courtroom. Indeed, it is difficult to imagine a more controlled presentation of “in your own words” narratives than testimony at a criminal trial. Judges constantly (and properly) rule that witnesses may testify about one set of observations but not another, on grounds of relevance, unfair prejudice to one side or another, or observance of the many rules that bar the admission of relevant evidence, including prior crimes and privileged information.

Though witnesses at a criminal trial have free speech rights as broad as those of anyone else when they are not testifying, in court, witnesses do not have constitutional protection that extends either to the content or to the form of their presentation.

That said, there is at least a policy reason to avoid forcing crime victims to phrase what they say in terms that may implicitly negate the truth of their own sworn testimony. It is already often intimidating and traumatic for witnesses to speak in open court about crimes that resulted in both physical and psychological damage. Though victims understand that a jury will decide whether or not to convict, they also expect to be able to describe what happened to them in words that accurately represent reality.

For a rape victim to be told, “You must not say that you were raped but that you and the defendant had intercourse” is to compel her to distort, rather than reveal , the painful reality of what occurred. Though she does not have a First Amendment right to tell her story, it seems unnecessarily cruel and degrading to require her to make her story sound more “neutral.” And in this context, there is no truly neutral word; most lay people would take someone saying “We had intercourse” to mean “We had consensual intercourse.”

There is still good reason to strive to protect criminal defendants from whatever has gone wrong in our courtrooms to generate so many convictions of innocent people. Perhaps we might provide expert witnesses to help cast doubt on the accuracy of eye-witnesses identifications, a frequent source of error in criminal cases that nonetheless represents the “gold standard” of proof in the eyes of many jurors.

We should not, however, reinstate a variant on the old opinion rule, in which we require prosecution witnesses to speak as though they must share the presumption of a defendant’s innocence with the jury. Such a rule misconstrues the presumption of innocence and simultaneously detracts from the coherence of a narrative that jurors seeking the truth have an interest in hearing. Word bans, finally, harm the witnesses who appear in court to testify against criminal defendants, sometimes at great personal cost. It does so by putting words in their mouths that minimize the gravity of what they have suffered. Though they mean well, courts should thus reconsider the wisdom of “word bans” at criminal trials.

Sherry F. Colb, a FindLaw columnist, is currently a Visiting Professor at Columbia Law School and will be joining the Cornell Law School faculty in the fall. Her book, When Sex Counts: Making Babies and Making Law, is currently available on Amazon.