Wednesday, November 17, 2021

Mistaking Inferences for Penalties

by Sherry F. Colb

Earlier this week, I listened to a debate on Intelligence Squared about whether Cancel Culture is toxic. As I knew would happen before the debate even began, the participants soon started arguing about whether Cancel Culture exists. Two believed that it does and two that it doesn't. If you think that Cancel Culture is a myth, then you are likely ignoring a lot of undisputed facts on the horizon (sorry). And if you believe that free expression means that no one should be able to draw inferences about you from what you say, then you are at war with the rules of logic. 

Of course there is such a thing as Cancel Culture. What do I mean by that? I mean that people who have said things that were either unpopular or insensitive or even offensive have faced consequences that were grossly disproportionate to what--if anything--they deserved. An atmosphere in which people of good faith routinely feel the need to say "I would never state this publicly, but...." is a Cancel Culture, and one cost of living in a Cancel Culture is that the apparently prevailing normative view of things might in fact bear little resemblance to the actual prevailing normative view of things. 

I will not specify examples of it here  because no matter what example I choose (and there are quite a few of them), I know that some readers will loudly condemn me for disagreeing with them about whether a particular person who voiced an unpopular/insensitive/offensive idea deserved to be treated like the second coming of Josef Goebbels or David Duke. Suffice it to say that disproportionate responses to relatively innocuous expression is easy to identify, and I'm happy to be more specific (and I have been more specific) with the small number of people I completely trust. Michael Dorf is one of the people in this trusted group, so he can confirm for skeptical readers that I am right and that trying to self-censor to avoid clear threats of demonization is exhausting and wastes energy that could be directed at trying to solve society's problems.

At the same time, however, when people say things, their statements give rise to logical inferences, and it is neither realistic nor fair to demand that nobody draw any inferences as a condition of free speech. Say John Doe makes the following statement to his friend Jason Roe: "You know women today are so demanding. It used to be that they cleaned the house, did the shopping, satisfied you sexually, and shut the fuck up.  Now everyone 'believes women,' and we have to apologize for having a dick. I really miss the good old days." Imagine that John Doe is a law professor and that some students overhear his comments.

No one is under an obligation to think that what John said is consistent with John being something other than a complete asshole. Was he free to say what he said? Sure. But people who heard it are also free to think he is a creep and a misogynist. Women are free to decide that despite his good looks, they will not be dating him any time soon. He is not entitled to people's good opinion. There is regrettably a shrinking space for permissible chit chat, but I would locate what Doe said as falling outside even a very capacious and generous space, one that we might have encountered ten years ago.

A case decided by the Eastern District of Pennsylvania, Bey-Cousin v. Powell, brings to mind the clash between penalizing expression in a Cancel Culture sort of way, on the one hand, and drawing rational and normatively proportionate inferences from what people say to factual and normative conclusions, on the other.

The case involves a man who had some trouble with the law and whose criminal conviction was reversed on appeal. The man brought a civil rights lawsuit claiming, among other things, that police planted a firearm on him. In defending the lawsuit, police wish to introduce into evidence the lyrics of some songs that the plaintiff wrote. Police believe the lyrics undermine the plaintiff's version of events and support the defendants' narrative.

The district court in the case ruled that the song lyrics would be excluded. I have no opinion on whether the particular lyrics should come into evidence or were properly excluded. In so ruling, however, the court invoked the First Amendment right to free expression in the arts. The court said that if an artist's lyrics were freely admissible against him in court, then that admissibility would have a chilling effect on artistic expression. In response to this constitutional concern, the court ruled that before being able to introduce an opponent's lyrics into evidence, the proponent must overcome a presumption that artistic expression is not factual. Furthermore, the court held, it is not enough to show a resemblance between events in the real world and the statements of the artist in his music or other artistic expression.

I have written about the difference between protecting the freedom of speech, on the one hand, and refusing to admit that speech as evidence of other wrongdoing in a court of law, on the other, here. An artist is free to write songs about his encounters with the law and with other people and to fictionalize those songs to his heart's content. But that freedom does not entail a right to suppress relevant evidence of wrongdoing in a lawsuit. He can, of course, argue that a particular piece of evidence, a song, is far more likely to prejudice the jury against him than it is to shed light on what he actually did. The district court did say that such a balance would favor exclusion in this case. But the court appeared to go further than that, asserting that in many cases, relevant evidence that could illuminate the facts for the jury would not come in, asserting that "starting with a presumption that artistic expression is not a factual admission might in some cases lead to the exclusion of admissible evidence. But the First Amendment requires no less."

The Supreme Court has had occasion to consider the question whether the First Amendment freedom of speech and expression ought to inform our understanding of other areas of law, even when no one is trying to chill or penalize speech or expression. In Zurcher v. Stanford Daily, the police searched a student newspaper office on the basis of a warrant and probable cause, and the newspaper claimed that because of the importance of freedom of the press, police should need more than the usual probable cause and a warrant that they must have to conduct searches of people's homes. The Supreme Court roundly rejected this claim and said that all searches (other than extraordinary ones like surgery to find a bullet as evidence) undergo the same analysis. So long as no one is deliberately interfering with the freedom of speech or the press, probable cause and a warrant are enough.

Many people (including me) are critical of the Court's Zurcher decision because allowing the sort of rummaging in a newsroom that police typically do when they search could really interfere with the freedom of the press, particularly at a time when everything was on paper that police could see during a search. A more demanding standard for searching a newspaper would therefore have made a lot of sense. But the Court said no. The First Amendment freedom of speech and expression does not change the operation of other legal regimes like the law of search and seizure.

In the context of free expression and the admissibility of evidence, there is even less reason to allow the first to alter the second than there was in Zurcher. Almost everything that goes into evidence in a court case consists of words or expression of one kind or another. If we were to take seriously the idea that the admissibility of speech imposes a presumptively impermissible chilling effect on free expression, then our generally liberal admissibility approach under the Federal Rules of Evidence would morph into a system in which all evidence would be presumptively inadmissible. This regime would completely undermine the principle that the law has a right to "everyman's [sic] evidence."

The court gives a few examples of fictional works of art that have no bearing on reality. One is "I Shot the Sheriff," written by Bob Marley and performed by Marley and later by anti-vaxxer extraordinaire Eric Clapton (who won't play to "discriminated [read "vaccinated"] audiences"). Neither Marley nor Clapton shot the sheriff, though Clapton may be responsible for more deaths for his COVID-promoting pontificating. Another example is Bohemian Rhapsody by Freddie Mercury and Queen, who were not confessing to mama that they killed a man. Of course, no one ever thought that these works of art corresponded to events in the real world, and their authors were probably not afraid that the songs might become evidence against them. 

Statements that people make outside of the artistic area are far more likely to become evidence against them than are song lyrics and therefore much more vulnerable to the chilling effect that the district court discusses. The district court, however, seems to focus exclusively on artistic expression rather than on all speech protected by the First Amendment, perhaps because it correctly understands that a system in which evidence comes in to prove facts requires access to speech and words as proof, regardless of how free that speech might be and how protected from punishment. Artistic expression is, of course, part of the freedom of speech, but the First Amendment does not single it out for special protection and indeed does not say anything about it at all. I am not an originalist, so I am not especially interested in whether people at the time of the framing and ratification of the Constitution or of the Fourteenth Amendment believed that art enjoyed the same sort of protection as other speech (though I suspect they did not). But art does not receive extra protection either.

The truth is that most art will never become evidence in court. If you write a novel about how much you hate some fictional person and how much you would like to kill him, it seems very unlikely that you will come to be accused of murder and then have your novel used in evidence against you. And even if that improbable event takes place, you can explain to the jury that fiction is often different from reality. The problem with unduly prejudicial evidence is that juries fail to understand why the evidence is not really all that relevant. But juries are pretty good at understanding that novels and songs are not the same thing as reality except when they are. On occasion, a song's lyrics may track what happened in a particular case quite well and therefore have a place in the jury's determinations about what happened. If John Doe of our earlier example is suddenly widowed by the murder of his wife, and he has, coincidentally, also written a song called "I Killed The Bitch," it does not burden the freedom of speech and expression unduly to allow the jury to take a look at his art.


Joe said...

Thanks for letting us know Prof. Dorf can be trusted. I was on the fence.


The linked case is helpfully only seven pages long.

I'm surprised the question appears so novel. Would think it would have arose numerous times at this point. There are so many cases out there.

The officers, like Prof. Colb, argues that the jury should judge the lyrics here. The judge was concerned that would "be a trial about an artist’s process, asking the jury to decide where the line is between inspiration and narration" instead of a search for truth. It would not be a good use of trial time given the rules of evidence.

The judge grants the lyrics can be used if the officers "offer some preliminary indicia that the artistic expression is a truthful narrative" but argues it was not used here. I don't know if the judge singles out artistic expression here except to argue it regularly is more fictionalized.

I am sympathetic to the judge's views here at least to some degree. A basis issue for me is that I'm really not sure how relevant the lyrics are. The case doesn't seem to put an absolute rule in place.

In a clearer case, I might think otherwise. But, I understand the other argument though given how often evidence is blocked from juries (maybe the professor thinks too much), I'm not sure how much this case really is much different. Some evidence deemed potentially misleading, mixed with other interests, is kept from the jury.

kotodama said...

[Vol. 1]

Joe, but she only said she personally trusts Prof. D., not that he could be trusted generally. So you may still have to be careful! Plus we have Prof. B. in the previous post being accused of putting words in people’s mouths. Trust concerns seem to abound on this blog lately! (/s)

Seriously though, I found the court’s analysis really underwhelming and shoddy. I don’t think it’d pass muster in an introductory Evidence course. So shocking this guy is a TP appointee. (Speaking of relevance, the fact of the TP appointment is always relevant IMHO, and I’d prefer it be noted upfront.)

For one, although Joe observes the decision PDF in its entirety reaches (close to) 7 pages, the actual core analysis (§ III.A) only covers somewhat over 3 pages. It’s even less if you discount the first paragraph of that section, because it basically just cites a few more legal standards. So the barebones nature (while the soul of wit no doubt) is sort of a red flag right there, considering the court’s making a decision that, while evidentiary in nature, also has an important Constitutional dimension to it.

Along the same lines, it’s concerning the court doesn’t cite any authority (aside from the elementary legal standard, which is nothing special) in support, when again, it’s effectively ruling on a Constitutional issue. It doesn’t even cite a law review article, let alone actual caselaw.

The court protests that “[t]he Parties have not identified, and the Court has not found, a case” that’s on point, but (1) that’s a lousy excuse and (2) I don’t even buy it. Just off the top of my head, I can think of some highly relevant—and famous—caselaw, as well as scholarship. For example, in the former category, while not precisely on point of course, Wisconsin v. Mitchell had some quite important things to say on the subject. One of those was: “The First Amendment … does not prohibit the evidentiary use of speech to establish the elements of a crime or to prove motive or intent. Evidence of a defendant’s previous declarations or statements is commonly admitted in criminal trials subject to evidentiary rules dealing with relevancy, reliability, and the like.” Surely a guy who graduated HLS, clerked for a federal judge, and practiced for almost 2 decades would know that one. And if the First Amendment poses no independent bar to the “evidentiary use of speech to establish the elements of a crime” even for defendants facing jail time, it beggars belief it would somehow prevent introducing a party’s own statements to undermine the elements of a claim especially when that party is the plaintiff in a civil case merely seeking damages. Oh, and Mitchell was a 9-0 decision, written by Rehnquist and joined by every liberal on the Court at the time. (I recognize the Court noted standard evidence rules still apply, but for now I’m just puncturing the sails of the First Amendment argument.)

Likewise, in the latter category of scholarship, there is a plethora of material to cite. Over at EvidenceProf Blog, South Carolina’s Prof. Miller often discusses cases where—like here—the admissibility of rap lyrics is at issue. Below is a recent example:

He’s published articles on the subject too, as have others (e.g. SSRN abstract IDs 2476337 & 3369175). Oh, and did I mention people have also addressed the topic already in a full-length book?

Of course, I’m not saying the court had to agree or disagree with any of the positions being taken by scholars. But at least acknowledge the scholarship exists. Otherwise, you’re not really doing the bare minimum your job requires—and that your credentials suggest you should be capable of.

Since this portion is edging up to the character limit and I’m about to segue to evidence issues anyway, I’ll continue in a separate Vol. 2.

kotodama said...

[Vol. 2]

Like the Constructional analysis—and Prof. C. had her own separate critique there, which I obviously won’t duplicate—the court’s treatment of the evidentiary issues was incoherent and lacked rigor.

At the outset, it’s not even clear which Rule is really being addressed. The court initially cites Rule 401(a), which goes to straight relevance. But then it mentions Rule 104(b) for the distinct issue of conditional relevance. And then to make things even more confusing, it states the question as “determin[ing] when artistic expression is relevant.” Since it never cites either Rule 401(a) or 104(b) after that—to say nothing of, you know, actual caselaw or other authority—the basic framework of the analysis gets quite muddled. On top of that, you have this made-from-whole-cloth “presumption that artistic expression is not factual” which fails to address both whether a piece of evidence is relevant and whether it’s conditionally relevant.

If, for the sake of argument, we assume it’s the conditional relevance analysis, I don’t think it works. Conditional relevance means that before evidence of A can be admitted, its relevance has to be established by demonstrating a separate fact B. An example is evidence that a homicide victim accused the defendant of serious sexual misconduct. That evidence is potentially relevant to the defendant’s motive, but only if the defendant knew of the accusation. So the defendant’s knowledge is the conditional fact that also has to be shown before the accusation evidence can be deemed relevant.

Here though, I didn’t see any indication the court was chastising the defendants for not demonstrating some other, distinct facts to make the lyrics relevant. Instead, the court simply concluded that the “Officers … have not put [forward] enough facts to rebut the presumption that Mr. Bey-Cousin’s lyrics are art.” So what though? Proving that something isn’t “art”—whatever the court even means by that—has little to do with producing facts that could make some piece of evidence relevant. In fact, something can be both art and relevant at the same time. They’re not exactly mutually exclusive.

On that note (no art pun!), the straight relevance analysis has similar faults. Again, whether or not the officers really proved the lyrics were “factual”—assuming that’s a meaningful question—the issue is pretty much beside the point when it comes to considering relevance. It’s also just bizarre as an evidentiary matter, especially when it comes to relevance. The threshold to surmount for showing relevance is absurdly low. Does anyone seriously think—under a proper analysis—the lyrics here aren’t at least relevant to the claims? Of course they are. I just don’t see how one can escape that conclusion.

Accepting the court’s pulled-from-thin-air “presumption” would be problematic for other reasons too. It would basically require the proponent of the evidence—i.e., the lyrics—to decisively prove the events recounted in the lyrics actually happened. But of course, that’s not how the evidence rules work at all. They don’t require someone to prove up the actual merits of a claim (or here a defense) in the limited context of an evidentiary ruling. Again, they’re just about getting relevance, nonprejudicial evidence in front of the jury (or judge as factfinder). It’s the factfinder’s job to, you know, find the facts and determine the merits of the case.

kotodama said...

[Vol. 3]

The court does also wave in the direction of Rule 403 at the end, but it’s a complete throwaway. That’s ironic, because the abundant caselaw and scholarship on introducing lyrics, especially rap lyrics, has spent considerable time analyzing the issue under Rule 403. One doesn’t have to take a position on the ultimate outcome to recognize that Rule 403 at least the proper starting point. I suspect the judge—as a TP appointee—didn’t find the prospect of a full-blown 403 analysis very appetizing because it would require addressing the prejudices some folks have against rap music and its creators, e.g., indigent minorities, especially African-American ones. A First Amendment ruling free from any considerations of racial discrimination might then be quite useful in the service of white reactionaries in future cases. I don’t think it’s hard to imagine potential scenarios in the latter category either.

I hope this case gets appealed so there’s a possibility of the flawed analysis being rejected, although who knows if that would actually happen. At CA3, I could easily see TP appointees like Bibas and others signing on to this absurd reasoning, maybe after sprucing it up somewhat to make it more presentable.

Joe said...

Yes, the breadth of our trust is an open question. Like the former Mets manager now has a more limited role in the Yankee organization. He might be better trusted doing that.

Wisconsin v. Mitchell is cited in the Verdict column linked in the OP. I'm not sure how much it is directly relevant to the issue of the lyrics being put into evidence. The opinion doesn't say such lyrics can never be submitted.

Anyway, the opinion (claims) that the "parties" didn't identify relevant cases regarding the rule of evidence cited. The citations by K. are appreciated though w/o seeing the briefing, I can't know if the Harvard grad judge here simply missed what was put in front of him. If a more in depth analysis on appeal refutes the article here, I'm quite open to reading it. My take is hedged.

As an aside, Judge Bibas is a friend of K's pal over at a certain sentencing blog.

kotodama said...

I appreciate the response from Joe.

I'll cop to not having read the cited Verdict column, but I think it just proves my point. While I agree Mitchell doesn't necessarily decide the exact question in play, it's clearly a relevant, high-profile case that should at least be cited—as everyone seems to recognize except, for some reason, the actual judge, who presumably fancies himself a pretty learned and smart guy.

Also, the judge didn't just blame it on the parties. He also said—as I quoted above—"the Court has not found[] a case" on point. Again, I don't think that meets even the most minimal due diligence standards for this kind of analysis. It wouldn't be any different if the judge had completely relied on the parties either. To employ a rap metaphor, you can't freestyle just because you feel like it. Especially here, where the relevant caselaw and scholarship is plentiful and not particularly inaccessible either. It's obvious he spent some time on the decision too, what with the silly Van Gogh quotes and all. He just did a lousy job when it comes to, you know, actual legal authority.

Joe said...

BTW, I appreciate the references to cancel culture and the usual verve of the author.

This seems to be a key point in the OP:

At the same time, however, when people say things, their statements give rise to logical inferences, and it is neither realistic nor fair to demand that nobody draw any inferences as a condition of free speech.

I think that is a fair rule. As I understand it, courts block many things that many might think warranted admission under it. I appreciate the reply to pushback on my comments. To quote the judge, they might be considered "inchoate."

Greg said...

In some sense, I think there may be a racial component here, and reasonable concern that what is typical in certain forms of modern music intended primarily for consumption by cultural peers will be misinterpreted by those who are not well-versed in that culture. It has the potential to extend the trial due to an extended discussion about musical tastes, which I can understand the judge wanting to avoid. Now, even if the judge wanted to avoid that, it doesn't mean the law entitles them to do so.

kotodama said...


"courts block many things that many might think warranted admission under it."

Yes, that's obviously a big aspect of the evidence rules. To be clear, I'm not complaining about that aspect as a general matter. I'm not even necessarily upset the lyrics at issue here were ultimately excluded. Maybe that was indeed the correct bottom-line result. I just take issue with the legal methodology, which isn't only hopelessly flawed, but could also have pernicious consequences (IMO, at least) down the road, if it gains traction anyway.


Agreed about the possible racial component. But if there is one, it both can and should be handled via Rule 403, IMO at least. As I've tried to explain above though, the reason 403 might exclude the lyrics has nothing to do with them being "art" or lacking relevance. It's quite plain the lyrics are relevant—regardless of whether they qualify as "art"; and surely they do—no matter how strenuously the court tried to avoid that inevitable conclusion.

Despite being relevant, 403 could still exclude the lyrics if that were outweighed by, e.g., unfair prejudice. I could easily see the potential for unfair prejudice based on racial bias when dealing with rap lyrics. Especially where—as here—the relevant party (Bey-Cousin) is African-American, admitting the lyrics might bring to the jury's (or a judge's, but hopefully that's unlikely) mind ugly stereotypes associated with rap like gangs, violence, drugs, etc. If so, the jury could very well find against Bey-Cousin simply because of their general aversion to rap music and its negative connotations, as well as people who make such music. A decision made on that basis would of course have little or nothing to do with a sober assessment of the actual evidence presented in the case, and so would be quintessential unfair prejudice.

Now, it's true Rule 403 recognizes other grounds for excluding otherwise relevant evidence. And it's also true the judge at least paid lip service to some of those grounds—undue delay and waste of time, which in my view are sort of equivalent—at the very end of his analysis. I believe that's what you meant by "the potential to extend the trial due to an extended discussion about musical tastes"?

In response to that concern being put forward, I'd have to seriously disagree that a "collateral inquiry into Mr. Bey-Cousin's artistic process"—as the judge put it—would be at all necessary. As intricate and thoughtful as his "artistic process" may well be, delving into that seems totally beside the point. If using the lyrics were permitted, all that would happen is this. Defendants' counsel would cross B-C about certain of the lyrics. In response and/or potentially during redirect as well, he could simply explain—if that's in fact the case—that the lyrics in question were fictional or greatly embellished, etc. Even a juror with just a high school degree knows quite well that song lyrics aren't exactly the same as a sworn witness affidavit.

What's described above happens on the regular whenever witnesses are confronted with their own unhelpful statements. They try to explain away the bad parts, and the jury ultimately decides what to adopt as the truth. It's not usually very time consuming, and it can happen with statements considered to be "art" as well as many other kinds of statements. There's just no reason to think in this particular case that B-C’s lyrics are a special exception to the ordinary scenario or that some extended "collateral inquiry"—especially into the minutiae of his "artistic process"—is needed to explain why the lyrics ought to be taken with a dose of NaCl.

Again, you don't have to take my word about any of this. I've already pointed out the abundant caselaw and scholarship addressing these issues. But this judge is apparently so full of himself he really thinks he's mapping out uncharted territory.

Greg said...

Rap music in particular seems to have a significant amount of what appears to be suggesting violence, but that really isn't actually about that. I expect to really convince a jury that rap music isn't, for instance, actually advocating killing police officers would require more than a simple "fiction isn't real."

In a sense, both white and black communities often have the attitude that cops are powerful individuals to be feared, but they way they express that in different ways, partially due to the level of risk involved, and the perceived likelihood of being able to improve outcomes by acting respectful. This results in cultural misunderstandings that, as a whole, make rap music more believable to a white juror as an indication that a black defendant is a threat to the community than a black juror would perceive that same music.

It's good to know that there are existing mechanisms like 403 that allow exclusion of prejudiced information, and that the concern here is HOW the evidence was excluded.

kotodama said...

Greg, I appreciate the productive discussion here. (Joe too, of course!)

Your first paragraph raises an interesting point. I think you can actually spin that out into two distinct scenarios.

(1) would be trying to introduce evidence simply to show that someone is a rapper or a big fan of rap music—but with no relevance at all to the particulars of the case. That's clearly improper under 403 every time because the probative value is always 0 while there's at least some or potentially a great amount of unfair prejudice.

(2) is the same as (1), but the evidence, e.g., rap lyrics, is relevant to the case. Then, you will have to balance that against the unfair prejudice—which still potentially exists, of course. But Rule 403 is very flexible and designed just for situations like that. On average, it should reach an appropriate result. And serious mistakes can be corrected via appeal.

I think B-C's case falls pretty clearly into category (2). But either way, 403 can and should handle it. And in either case, the risk of unfair prejudice wouldn't arise simply because he's engaging in "art" or any other kind of speech that might have some fictionalized elements, but from a very specific connection between one kind of art/speech—rap—racial prejudice.

The same would hold generally true I think even for a lily white rapper. But I'd expect the 403 analysis would differ somewhat because the potential unfair prejudice might be less in that case.

At risk of piling on with this judge yet again, the bottom line is he fundamentally misunderstands what relevance is all about in the evidence realm. Making a proponent prove upfront that some evidence is 100% true as part of the initial relevance inquiry gets things completely backwards. The whole point of factfinding in a trial is that deciding what evidence to believe—i.e., whether something's fact or fiction—is the sole province of the jury (or judge in a bench trial). Relevance is only a threshold inquiry. It doesn't concern itself with evidence being true or not precisely because that's what a jury will figure out later. The only point of relevance is whether evidence—true or not—even has a connection (lawyers love to use the word "nexus" too) with a particular case or not.

I hate to belabor the point, but let me close with a very simple example. Under the judge's logic, a witness might not be allowed to testify because we don't know beforehand if the witness will tell lies (fiction) or not (fact). So, the only way you can get witness testimony at trial is if you prove upfront that all the testimony will be truthful. In other words, the jury would have nothing left to do in terms of evaluating the witness's credibility. That's obviously absurd. But it's what the judge's logic seems to demand.