Free Speech and Criminal Intent

Listen & Download

Terroristic threats straddle the line between criminal acts and constitutionally protected speech. This is especially so when the apparent threats could also be interpreted as "artistic" speech, say, as song lyrics. How can law enforcement and courts tease out "true" threats from, say, outlaw country music or hip-hop? When should society trade the right to hear the next "I Shot the Sheriff" for protection against illegal intimidation? 

Master attorneys John Elwood and Michael Dreeben litigated these questions from opposite sides at the U.S. Supreme Court. Listen to them continue the conversation this Friday at noon, with moderator Prof. Eugene Volokh - America's most prominent scholar of speech and the law.

Event Transcript

Operator:  Welcome to The Federalist Society's Practice Group Podcast. The following podcast, hosted by The Federalist Society's Criminal Law & Procedure Practice Group, was recorded on Friday, September 13, 2019, during a live teleforum conference call held exclusively for Federalist Society members.


Wesley Hodges:  Welcome to The Federalist Society’s teleforum conference call. This afternoon’s topic is on “Free Speech and Criminal Intent.”  My name is Wesley Hodges, and I’m the Associate Director of Practice Groups at The Federalist Society. 


      As always, please note that all expressions of opinion are those of the experts on today’s call.


      Today we are very fortunate to have with us three premier experts on this topic. And first today is Mr. Michael Dreeben, who is the former Deputy Solicitor General of the United States for the Department of Justice. Also with us is Mr. John Elwood, who is a partner at Arnold & Porter. And our moderator today is Professor Eugene Volokh, who is the Gary T. Schwartz Distinguished Professor of Law at the UCLA School of Law. After our speakers have their back and forth and remarks for the beginning part of this call, we will have an audience Q&A, so please keep in mind what questions you have for this topic. Thank you very much for sharing with us today. Professor Volokh, I believe the floor is yours.


Prof. Eugene Volokh:  Great. Thank you very much. Always a great pleasure to do one of these teleforums, or is it telefora? We’re talking today about threats and about the true threats exception to the First Amendment. There’s broad consensus that there ought to be such an exception. The Supreme Court has long held that there ought to be such an exception. Interestingly, the first threats cases were actually threats of employment retaliation in the 1940s, and the question was whether those could be punishable when employers engage in them or perhaps even unions engage in them. And the Court was willing to say yes, that they ought to be at least civilly actionable. And later on, the Court made clear that threats of criminal conduct could themselves be made criminal.


The question is how broadly does that exception reach? And there are at least three dimensions—there are actually more—but at least three dimensions to this question. One is what mental state is required? What mens rea is required on the part of the speaker in order to make something a punishable threat? So one possibility would be to say that speech is a punishable threat only if the speaker has the purpose, the specific purpose of putting someone in fear. That would be perhaps consistent with the way incitement is treated, where there’s a purpose element to insight. Although, incitement, of course, that’s advocacy of violence, but you could imagine something similar for threats of violence.


Another possibility is that it’s enough if the speaker knows that the audience is going to perceive it as a threat. A third possibility might be it’s enough if the speaker is reckless about the possibility that the audience will perceive it as threat. That kind of recklessness or knowledge standard is kind of similar to the actual malice standard as to truth or false in liable cases. Another possibility might be that it’s enough if this speaker is negligent about the risk that the audience will perceive it as a threat. So it’s enough if a reasonable listener would perceive the statement as a threat. You could imagine, of course, strict liability, as well, but I don’t think that’s been really seriously discussed, in part because if somebody wouldn’t reasonably perceive that the audience would perceive it as a threat, we usually don’t call it a threat. If I say something and it turns out someone for some reason, that I could not reasonably anticipate, views it as threatening, that’s just not something we’d normally understand as a threat on my part.


So that’s something that’s drawn a lot of attention of a lot of courts. There was a split on the subject amongst various state and federal courts. Interestingly, I think, as a practical matter, it probably doesn’t matter that much in a typical case. It matters a lot as to how the jury’s instructed. But I think in a typical threats case, the jury would reach the same result, whether it’s conviction or acquittal, regardless of what mens rea they’re asked to find. Although, as we probably will discuss later on, there may be a few cases in which the required mental state is important.


Note that the mental state is as to how a listener would perceive the statement, whether the listener would perceive it as a threat of violence or a threat of illegal conduct. It doesn’t have to do with whether the speaker intends to carry out this threat. A lot of punishable threats are threats that are bluffs but that are perhaps not perceived or intended not to be perceived as bluffs on the part of the speaker.


So that’s the first question, the mental state required. The second question is how do you tell a true threat of violence -- by the way, the exception is often called the true threats exception. How do you tell a true threat from a joke or from political hyperbole or from fiction or art? That’s something that courts have also been struggling with just in part because the meaning of words often is quite different in different context. By the way, a similar issue, although not identical, arises in lots of other areas of free speech law, such as libel. One question that often arises in libel cases -- how do you tell something that is a factual assertion that could be libelous from hyperbole or from figurative language or from parody, let’s say, or fiction? So this question comes up in threats, as well. Although, because threats are generally criminally punishable, there’s a lot at stake there.


There’s a third question. How does one distinguish threats from warnings? Let’s say, for example, somebody says, “If some police officer comes and attacks -- and tries to arrest me, I’ll kill him.” That would be a punishable threat, especially if it names a particular police officer, but maybe even if it says, “if the city police do that.” On the other hand, imagine somebody says, “You know, I just want to tell you, if the police keep being abusive the way that they have been, somebody’s going to take a shot at them.” At some level, that’s got to be permissible and perhaps even a friendly warning in some situations. We should warn you this kind of law is going to lead to violent resistance, let’s say. Sometimes it might be not so friendly. In fact, often, it might be not so friendly, but it would, I think, be permissible debate about what the police ought to be doing. But at times, it’s pretty clear that this is a camouflaged attempt to make a direct personal threat through the language of warnings.


But how does one distinguish the two? That ends up being very difficult. There are other questions involving blackmail and the blackmail paradox and other things, but we’ll probably avoid those. But these three, I thought, might be worth discussing, especially the first two. And with that, I turn things over to our panelists.


John Elwood:  As far as mens rea goes, I’d be interested to know what Michael’s take is because, as the former Deputy Solicitor General who oversaw, essentially, the entire criminal docket at the Justice Department, he may have his own take on how much mens rea matters. But the current state of the law, at least the last I’ve paid attention to it, was that much of the federal courts had adopted essentially a negligence standard. That is, if you said something and a reasonable listener would interpret it as a threat, that was enough to be convicted of violating the federal threat statute, 18 U.S.C § 875(c).


As an aside, a great many states have adopted a purpose standard -- I’m sorry, an intent standard. I don’t think they really say that much beyond that, and intent is a very malleable word. But many of the big states have adopted intent-to-threaten standards for their own stat threat statutes. California, Texas, New York, Illinois, Pennsylvania, Ohio, Georgia, Indiana, Massachusetts, and then Colorado are the ones I have written down.


But at the federal level, only the Ninth Circuit had adopted an intent standard. Although, again, intent can be a little bit murky about whether it’s a purpose or whether you merely know that it’s going to happen and go ahead and do it anyway. And that’s where the law was before the Elonis case, which was argued in 2014 and decided in 2015, where the Supreme Court at least crossed off one thing, which is that they struck down the negligence standard. They said you had to have some sort of mental state, but they didn’t go beyond that and answer that further question. I don’t think that there’s been a lot of development in the law at the federal level since then, but perhaps Michael knows more.


Michael Dreeben:  Thank you and thanks for the setup, Eugene, and, John, for your bringing Elonis into the discussion. I think the first thing to observe here is that there are two layers of the analysis. One layer is statutory construction. The second question and layer is what does the First Amendment require. And in the federal system, the Court settled in Elonis that negligence is not enough. We, as the United States in that case, did not proport to characterize the prevailing standard as the negligence standard. But if you plug in the general definition of negligence, which is you act in a way that’s contrary to what a reasonable person would do under the circumstances, then it has that dimension of negligence. Everybody agrees that somebody has to know and understand what words they are speaking.


The next question is, as a matter of statutory construction, congressional intent or state legislative intent, what more is required to render a person criminally culpable. And the principle that the Court uses for statutory construction in federal cases is when Congress does not specify a mens rea, it is assumed that Congress does indeed intend that there be some layer of mental state to render a person culpable. And the test is when a person would be able to know that they are engaged in wrongful conduct as apposed to innocent conduct. And the minimum standard that achieves that goal is what the law requires.


So in this case, if you build off of the model penal code hierarchy, which Eugene summarized, you have purpose to achieve a result. You have knowledge that the result will occur, which basically means that you are practically certain that it will happen. Absolute certainty is not required. And then, recklessness, in my view, is very close. You consciously disregard a substantial and unjustifiable risk. That’s the model penal code phraseology. And I think I would just echo what Eugene said. It is really difficult to imagine a jury confronted with a guy who, say, is threating his -- well, I won’t use a characterization -- a guy who is saying his former spouse better watch out because, if he finds her on the street and sees her with another guy, blood’s going to run. And I think it would be tough for a defense lawyer to stand up and say, “He may have known the risk, and he didn’t really pay a lot of attention to it. But he wasn’t certain that it was going to occur, so you should acquit him.” It doesn’t seem like a very appealing jury argument. It might have more appeal at an appellate level in some cases.


And then when you get to the First Amendment question -- and there are some states that still adhere to the negligence standard so the First Amendment question will arise in state case. Although, I doubt it will arise in federal cases. Then the question is why do you need more than recklessness in order to afford adequate breathing space to free expression? That’s the question that I would put out there. I think it’s worthy of discussion. It may have some interaction with what you think it takes to make words objectively threatening, as opposed to a joke, hyperbole, satire, art, or the like. But in my view, the debate about mens rea, it will get settled that knowledge or recklessness probably doesn’t matter a lot. And once you reach that floor, I would question whether there’s any serious First Amendment issue left to decide.


John Elwood:  I think there’s a lot to that, and this isn’t necessarily stuff that needs a huge amount of breathing room. I think that there are instances, but the reason why I think Michael and I were brought on for this is that we argued the Elonis case against each other. And one point that I kept running up against during the oral argument is the idea that this is kind of low value speech, and so why do you need much breathing room around it? I think there can be cases where it’s harder to tell if it was a threat. I think some of the better examples of them all involve people making what may or may not be tasteless jokes about pressure cooker bombs and things like that, which is something that I encountered when I was preparing for the case, examples that would come up from people who had actually had things happen to them. But query where you need kind of the breathing room on that. And you can kind of hypothesis them, instances where you might need some breathing room.


But a lot of the stuff, like the statement that you gave -- “Blood’s gonna run if I see you on the street,” or something like that -- it’s not the sort of thing that’s terribly sympathetic, and you don’t get a lot of room for it. In order to try to argue for a stronger mens rea, you really have to hypothesis something more akin to very strong talk and sort of the political atmosphere or something like that, or political milieu, in order to make it seem like this is something that needs more breathing room. Even though it was kind of contrary to my client’s position in Elonis, I think it is true. Even recklessness, if you’re consciously disregarding a risk that will be perceived, gives you a fair amount of breathing room, I would think.


Michael Dreeben:  And just to amplify that, it’s not just consciously disregarding a risk if you use the model penal code formulation. It takes substantial and unjustifiable risk. So that would be, I think, one quarter, perhaps, of the answer to Eugene’s third issue of distinguishing warnings from threats. There’s another answer that I tried to give in the oral argument in Elonis, which I’ll circle back to. But I wanted to back up a little bit because I think that, in trying to consider these issues, both what kind of mental state do we want before we’re going to punish people and what gives them adequate breathing space and what is going to be culpable and sanctionable as an objective threat, it’s worth stepping back to examine the reasons why true threats are outside of the First Amendment to begin with.


And that reasoning is that they engender fear. They cause disruption because people react to the fear. And then that gets balanced against the value of the speech, whether there are other ways to convey the message and so forth. And once you kind of look at it through the lens of its ultimate purposes, I think it brings some clarity to the objective manifestation that trigger threat liability. One issue that I think repeatedly comes up is specificity of the target. When you have, as in Elonis, a guy who’s making statements specifically about his wife and he knows that his wife is reading them and he’s been given a protective order to, say, stay away from this person, stop harassing them, there’s a lot better reason to infer that what’s he’s trying to do is instill emotional harm in another specific human being.


And I realize we’re not talking here about purpose. But it is certainly kind of a bedrock societal view that people who are trying to achieve bad ends are more culpable and more sanctionable than those who may happen upon them. I think the specificity of it helps you distinguish between is this the kind of thing that’s a threat, or is it going to be entertainment, political speech in particularly inflammatory terms, some other kind of response that’s maybe a joke?


The second thing that I wanted to say is, to bring jurisprudence into this, the Court did have a cert petition filed a couple of years ago in a case called Perez vs. Florida, where Florida was adhering to the negligence type standard of it’s enough that you know that you utter words. And it’s enough that you know -- it’s enough that a reasonable listener would interpret them as a threat. You don’t have to appreciate that yourself. It’s just you know you’ve uttered the words, and they are in fact what are reasonable listener would interpret as a serious expression of an intent to do harm.


The Court confronted that issue, denied the petition, but Justice Sotomayor wrote a concurrence in the denial saying, “Look. The issue wasn’t properly raised below, but I think that this raises a serious First Amendment issue.” And she cataloged the fact that courts disagreed with each other on that issue. Virginia v. Black, a cross burning case, has given rise to some confusion about what the proper mens rea needs to be as a matter of First Amendment law. And she thought that that case kind of exemplified the problem.


It was a guy who was very drunk. He walked into a bar. He was talking to his friends about a molly cocktail, which apparently is some kind of a drink, and it was misinterpreted by the bartender as a Molotov cocktail. And then because Mr. Perez probably had a few to many mollies himself, he came back in and he said, “Yeah. I’ve got one Molotov cocktail, and it will blow the whole place up.” And then he came back later and said, “I’m going to blow this whole, expletive deleted, world up.” He got prosecuted and convicted, and it kind of raised the difficult issue. And I would put it back to John. Is it enough to trust juries with a definition of what a true threat is articulated the way that Eugene articulated it? Is it enough protection to trust juries to sort the joke, the hyperbole, the satire from the serious expression of an intent to do harm?


John Elwood:  Interesting. Because this is also something that can be done by a judge as well because, after all, the case that gave rise to the true threat was something where the Supreme Court just said this is not a serious expression and so forth. And there was another case involving an overenthusiastic sports fan in St. Louis, I believe, who, again, they said as a matter of law it did not rise to a true threat. I think that there is some risk, particularly kind of at a negligence level, because, as I said during the Elonis argument, there is kind of a bootstrapping quality at the negligence level. Because the fact that it’s being prosecuted means that police, who are presumably reasonable, thought that it was a threat. That, I think, at the negligence level may not give all that much room.


You definitely see, though, and I sort of feel like there’s a wide variation in how carefully courts enforce this sort of true threats standard because -- like, for example, the case that Eugene flagged for us, which is State of Missouri v. Metzinger, involved a person talking about he’d posted tweets about pressure cookers around the World Series games. And the court there said that those were -- said basically that sports rivalries are famous for getting people’s blood up, so they couldn’t really be interpreted as serious expressions. Nobody would reasonably interpret them as serious expressions of an intent to do harm. That was not an outcome that I think was completely foreseeable. You couldn’t say that with any degree of confidence, and I’m not sure that a lot of other courts would agree.


But that said, you do wonder how much breathing room do you need for pressure cooker jokes, as opposed to I think that there’s kind of more need for room in sort of the political sphere. And one of the things I tried to do in Elonis was wrap myself in the mantle of not a guy threatening his ex-wife, but we have to develop a rule that is useable for people posting about the police reaction to demonstrations in Ferguson, Missouri a few years back or people using gun site logos on opponents of gun control and things like that, where it may be an infelicitous thing but you run the risk of chilling potentially political speech.


And to use an example from just last night, Beto O’Rourke said in the Democratic Debate, “Hell, yes, we’re going to take your AR-15.” Within minutes, a Texas state representative, Briscoe Cain, had posted, “My AR is ready for you, Robert Francis,” where there’s definitely a difference of opinion about whether that is a true threat. Although, I think that in the context, you would have to say that it is not. It is something that I would feel more comfortable trusting juries to make that decision somewhere above the negligence level of mens rea.


Michael Dreeben:  John’s bringing in of specific examples is extremely helpful because, when we talk about this abstractly, it either can seem too easy or seem intractably hard. But it definitely seems abstract. So I think two examples that are worth framing the issue here are one from the Supreme Court’s Watts decision and then one that’s much more recent. In Watts, you had a demonstrator during the Vietnam War who was speaking to an audience, and he basically said, “Hell, no. They’re not going to draft me. I’m going to burn the draft card.” People cheered and said right on. And he said, “And if they do draft me and give me a gun, the first person I want to get in my sights is LBJ.”


And I can tell you from experience that the Secret Service takes incredibly seriously any possible threat in language that’s directed towards the president. One of the sort of hazing exercises that used to occur in the Solicitor General’s Office is the Secret Service would go out, and it would find some guy who was in a mental institution. And he had written some letter on the back of an envelope that was barely understandable but conveyed something like, “I’m going to get the president and burn him down.” And he would get prosecuted and convicted. And the person who assigns recent opposition to young assistants would hand those things out just as a kind of test to see whether you would come in and go, “I don’t really know whether this actually makes any sense.”


So the Secret Service is very aggressive, very protective. And as John said, the Supreme Court got the case and, rather than resolve the more difficult mens rea issue, said this in context, given the audience and given the audience’s reaction, should have been understood as political joke, satire, and hyperbole.


And then one factual instance that is from a more modern situation is the facts of Knox v. Pennsylvania, which was the subject of a certiorari petition that the Court denied just last year, asking it to explore whether a subjective standard alone, regardless of objective characteristics of a statement, was sufficient to make it a threat consistent with the First Amendment. Whether that was a correct reading of the Knox opinion or not is a more difficult question, but that’s the way the petition framed it.


In Knox, you had some guys who were arrested by the police, and then they created a rap video, rap music, and imagery that picked up on the famous, or perhaps notorious, N.W.A. song from a decade or so ago about doing something to the police. And because we’re in a family audience, I’ll leave it to your imagination of what it is they were going to do. And in homage to this song, they included that lyric, but they also included a lot of specifics about the officers who actually arrested them and a lot of language which, if taken literally, was unquestionably threats to kill and kind of detailed about it, including knowing when one of the officers got of duty and where he lived.


They were prosecuted, and I think this is also responsive to something that John said. Police take this stuff very seriously when it’s directed at them. They have to go home at night, and we want them to be able to go home at night and lay down and go to sleep. They got prosecuted, convicted, appealed. It was upheld through the courts. And it tees up the question to what extent can you engage in vary vivid, violent imagery in a form that is unquestionable artistic when you know that, among your audience, not only in your audience, but among your audience are specific individuals who are named and who are going to regard it reasonably as a personal threat against them?


John Elwood:  And it’s interesting because what Michael said earlier about the specificity of target, it makes a huge difference because when you compare the Beatles song, “Run for your life. I’d rather see you dead, little girl, than to see you with another man,” just referring generically to a little girl, it’s a very different thing than when you’re basically mentioning police officers by name, which you really cannot recommend doing. It’s interesting though because I was surprised in the Knox case that they actually went to the trouble of recording a song. Admittedly, I wouldn’t say it had high production values or anything. But they went to the trouble of recording a song, which I think generally would be a stronger indication than just scrawling stuff that vaguely rhymes and posting it on Facebook -- that it was meant as an artistic work and not just trying to convey a threat.


But there are other cases on the other end of the spectrum. I guess it’s kind of the same end of the spectrum, but the context is different -- where the rapper Eminem, who Anthony Elonis in the case purported to look up to or he was trying to emulate, who wrote some very, very graphic and violent songs about his ex-wife. She was identified by name. One of them was even called “Kim” after her. It’s an interesting thing to think about that if he were an unknown and he had posted those lyrics on Facebook it could have very easily gotten him arrested. In view of the Knox case, even if he had just recorded them as an unknown, there might a case. Although, again, I can’t emphasis enough that things really started getting bad for Elonis when he started doing posts about the FBI officers who came to visit him. Context is something that matters tremendously, and it’s very hard to identify all of the various parts of context that matter.


Michael Dreeben:  What I think makes this issue so difficult in the internet age is lack of control or total foreseeability about the audience. What really emerges from the Watts case and the Knox case and Elonis itself is who your audience is really matters to the context in which you’re going to interpret it. If John and I had gone out after the argument and we’d gotten a beer and we were just kind of one upping each other with more outrageous threats just to have fun, even though the words on the page would be highly damaging to somebody’s wellbeing if they thought they were real, everybody who was familiar with the context would say they’re just having a good time. What if you’re doing that or what if you’re rehearsing for a play or doing something that you think is emulating N.W.A. and it’s overheard by others?


And in the physical world, you could say mens rea standard kind of takes care of that because if you don’t know that somebody is in the audience you obviously are not going to be able to be disregarding a risk that they will be offended by it. Is the same true on Facebook? This was something that the Court, showing remarkable prescience about the way that the internet works -- a little surprising in some respects, equally surprising that the Chief Justice being conversant with Eminem. But the question of, well, who is your audience if you post on Facebook; is it only your friends? Do you have to make a special effort to keep your page private? What is part of it is public, part of it is not?


You think the only people who were going to ever stumble upon your page are your mom and your high school friends, but it turns out that it’s actually broadcast, I’m using that metaphorically, in a way that people will see it in a community that just had a school shooting and will take it very differently. To what extent are you accountable for the wide variety of people who may be in your virtual audience on a digital platform? And John, I suspect you must have thought a lot about this because the statements in Elonis were on Facebook. I wonder if you have a view.


John Elwood:  Yeah. That’s interesting because I think even though -- I think with Elonis, even though his wife wasn’t his follower, I think it would be a pretty easy argument to make to the jury that he knew very well that she was in his audience. And I think it’s a very different case if it’s just a few guys or somebody with like four Twitter followers with no particular notoriety who says stuff for his audience of four and said something that then comes up in a word search of Twitter and somebody who is not among his followers takes it as a threat. So I think it can be a very plastic thing. But unfortunately, I think it would be an argument that you’d probably wind up making frequently to the jury.


But as an indication of how much your audience matters, one of Anthony Elonis’ postings was essentially he said, “That’s it. I’ve had enough. There’s enough kindergartens in ten mile radius of here that I could initiate the worst school shooting known to man.” And he posts this, and, at a minimum, it’s incredibly tasteless. And one of his Facebook friends actually liked it. If you see the evidence -- or, I’m sorry, the exhibit with a little exhibit sticker in the corner, there’s the little cartoon-y thumbs up posted by one of his followers right beneath it. So obviously, audience matters a whole lot. An important thing to remember is that the audience that you are directing it to may very well not be the jury veneer that they’re going to be drawing the jury from.


Prof. Eugene Volokh:  So can I chime in a probe a little bit further on this? I think these have been excellent, excellent examples and an excellent observation that, when you’re speaking in public, you’re speaking to an audience that you may be aware is very mixed. There may be some people in it that share your sense of humor. You may know there’s some people who don’t. Maybe some people have emailed you before, have responded before saying, “That’s outrageous. It makes me feel unsafe,” whatever else.


So imagine, for example, someone is posting something sharply anti-abortion and is saying, “Abortion is murder. And I’m just shocked all these people are willing to becalm all these abortionists, and therefore murders, especially since we’ve seen what’s happened to them. Some of them have been shot, and it’s right that they were shot because they were shot in proper defense of others. And I would think that people would think twice and ought to think twice before doing that, given that we think that they deserve to die,” let’s say, something like that. Not my views and generally not the way I would express things even if they were my views, but it seems to me that that is something that is within the proper bounds of ideological discourse.


But let’s say that this person, the speaker, is actually aware that some people will perceive this as threatening. And in fact, maybe there have been instances in the past where he’s posted similar things, and someone has complained to the police. So he’s aware that some people do perceive it as threatening, especially people who work at abortion clinic themselves and therefore have reason to worry about such things. So is that sufficient for recklessness purposes to get this man criminally prosecuted? Now, one possibility that was flagged, Michael, I think by you is that there’s got to be a substantial and unjustifiable risk. But I take it that a jury can’t say, “Well, this is a justifiable risk because this is really kind of important valuable political discourse. Whereas, if you were saying something about how police officers deserve to die, well, that’s an unjustifiable risk because there’s really nothing valuable about that.” Surely a jury can’t do that.


But if they can’t do that, then how do they determine whether this person was facing a substantial and unjustifiable risk or perhaps even knew that some people would perceive it as threatening and therefore should be punish or, on the other hand, that he was engaging in political discourse and shouldn’t be punished? How does one draw the line there, especially under a recklessness standard, given that you might know that at least some people in your audience do indeed feel threatened by such things?


Michael Dreeben:  I don’t think in those circumstances the mens rea layer of analysis does very much protective work for precisely the reasons that Eugene was alluding to. If someone makes you aware that they regard your words as threatening and you disregard that or override it and say it anyway because you think you’re entitled to do it as a matter of political freedom, the mens rea part of the analysis will not be hard. You’ve been put on notice. What will still remain is whether the statements objectively justify the label of a threat. And then there, of course, is a question of whether a jurisdiction could, as the petitioner in the Knox case claimed Pennsylvania had, do away with the objective test all together and simply punish the bad intent to threaten, even if it objectively did not constitute a threat.


That kind of analysis seems to me to raise much more fraught First Amendment issues because it would deprive society of the benefit of potentially, arguably useful political speech or other kinds of speech that reasonable people would understand are not intended to be serious expressions of an intent to do harm. One thing to sort of lay out a little bit in the argument on the other side of the position that both Eugene and I endorsed which is the recklessness knowledge line may not do a lot of work. Where I think may actually do some work is people who are drunk and people who don’t understand how social media work.


From my review of cases when I used to do this at the Justice Department, a lot of the people who are prosecuted for threats have had adult beverages that helped get them to the point of uttering things that they might not have in another circumstance. Some of them are just out and out mean people and racists, and they got what the wanted to do. But other people just got riled up. And I think that if you were sufficiently intoxicated, a recklessness standard you might not beat because people might not be so sympathetic with the fact that you got yourself drunk to the point where your inhibitions were down and you did something a little bit nasty. But if there had to be knowledge, you might be more successful in persuading a jury that “I didn’t know what I was doing. After 16 vodka tonics, it was lights out for me.”


And then on the question of social media, where the nature of the audience on Facebook does require some knowledge on the user’s part in order to have an understanding of it, I personally don’t think I have a really good understanding of who might be looking at a Facebook page. And that’s why I don’t post anything. I have now idea who’s actually going to be reading it. Would that be a defense if someone posted something on Facebook intending it to be read by friends and really maybe he had some shadow awareness that other people could read it? But he tries to say to a jury, “I didn’t have any idea that anybody else could do it. I never went through those inscrutable 17 paragraphs of boilerplate on the terms of service to get to the privacy policy where I could disable the control”?


Prof. Eugene Volokh:  That makes a lot of sense. Let me probe a little further. So one argument that people have given in favor of a purpose mens rea, again, a pretty high standard but one that criminal law sometimes provides and that the incitement test generally requires us to as to advocacy of violence—although, that’s not the only thing it requires—is that the purpose standard can at least provide some extra protection, however unpredictable it might be, to people like the guy I was talking about who is trying to kind of engage in political denunciation. Maybe some judge or jury will say, “Objectively, I think it’s a threat.”


But at least then the guy would have an opportunity to say, “Well, you don’t have enough evidence to prove beyond a reasonable doubt that I actually intended it as a threat. The mere fact that I happen to know that a few people would perceive it as a threat doesn’t show that I had the purpose of doing that, given that I know that lots of other people don’t perceive it as a threat and that my claim is my purpose was to speak to those other people and not to the few outliers who find it threatening.” So that’s one question that I had.


The second question, which reflects partly the fact that I’m actually not fond of purpose standards -- I don’t think they’re sufficiently protective, and sometimes I think they’re too protective in certain situations. But the second question that I had is how would be draw these lines between, objectively, what is something that should be seen as a true threat, apart from mens rea, from something that should be seen as political advocacy or as artistic expression or something like that? What is it, in the face especially of someone complaining saying “I found it threatening,” how is a court to determine?


And I know courts try to do this, but what’s the best way for them to try to determine whether something really ought to be understood as a threat for legal purposes and whether it ought to be understood as just politically hyperbole or not even hyperbole but kind of warnings or sharp denunciations?


Michael Dreeben:  I could offer a few words on that, and then I’ll turn it back over to John. Although, I do have some thoughts on the purpose standard as well. There’s a case in the Second Circuit called Kellner, I believe, that tried to grapple with some of the things that Eugene was talking about, although not all of them. And I think that it involved, if I’m not misremembering it, somebody who made a conditional threat that wasn’t temporally localized. And the language of it was definitely of the sort that, if taken literally, you would say, “Gosh, if I go to New York, I’m going to have to contend with someone who just said, ‘If he comes to New York, he’ll be taken out because he’s a nasty character.’”


And the Second Circuit, I think, focused on specificity, unconditionality, and temporal immediacy. So a little bit cutting close to sort of the incitement factors, as well -- may not do quite as good a job of segregating out the political hyperbole or joke context. So I think you have to add in one other ingredient, which a rational jury and perhaps a court applying de nova review would consider, and that is audience reaction. Because I do think it matters a lot if the audience laughs, as they did in the Watts case. Nobody was thinking that the speaker was really going to get LBJ in his sights. And if the audience is a narrower audience and people in the audience hear the guy’s threatening to bomb a kindergarten, I’m calling the FBI, it is some indication that maybe society doesn’t really take this -- this goes further than art.


If you took the entire speech and you put it in a play and it’s performed in a theater and people are in the audience, everybody knows that it’s entertainment. But if you take it out of that context and you want to play with sort of a reality television version of my art is going to have extra emotion infused into it because it treads the borderline between real and fantasy and that’s my whole point, then you’re running a greater risk.


Prof. Eugene Volokh:  Got it. So let me probe just a little bit more. I like the idea of trying to constrain the exception, but I wonder if sometimes that exception would be to -- the Kellner ruling would be too constraining and therefore might actually not be honestly followed by courts. So let’s take one example: conditionality. I take it if I were to say, “If you go out with my ex-girlfriend, I’ll kill you,” that’s conditional. But I should think that it ought to be punishable as a true threat, even though it’s conditional. Likewise, if I were to say, “Oh, I’m going to kill you in three months when I come back to the U.S.,” and you be let’s say ex-girlfriend or whatever else, or “One of these days, I’m going to kill you,” that doesn’t seem to have an element of immediacy.


But again, I would think that most courts would say that is prosecutable. Now, maybe they’re wrong. Maybe in fact they ought to be a requirement of unconditionality and immediacy. And if it’s a long term threat or conditional threat, well, nope. Wait until the person either tries to do something or makes a more immediate unconditional threat. But the alternative is to say, well, in fact for at least most kinds of threats, they ought to be punishable even if there is some sort of condition, which may actually make it more dangerous because it’s coercive and even if it talks about possible violence at some unspecified future time.


And maybe there ought to be some other mechanism or screening out the political threats that we think, or hyperbolic threats or some such, that ought to be protective. What do you think about that? Or do you think that ultimately conditionality -- unconditionality and immediacy are indeed the right tests?


Michael Dreeben:  I personally do not think they are for many of the reasons that you have articulated. It really goes back to what are the dangers of threats that we’re trying to eliminate. Compare it to, for example, to incitement. In incitement, somebody is advocating lawless action. That is not enough, under Brandenburg v. Ohio, to constitute punishable incitement. It has to actually have the eminent likelihood of succeeding. Why is that? Because the harm that incitement is trying to deal with is that explosive language will erupt into actual violence. Whereas in threats, the concern is that people’s lives will be disrupted, and they will have the emotional burden of feeling threatened. And a conditional threat can actually impose that in a greater way.


This came up in the cross burning case, Virginia v. Black, where some of the justices were probing on whether the burning of a cross should only be punishable if there were an eminent risk that it would lead to violence by the Klan. And the analogy that I tried to draw was that in the classic example of a sword of Damocles, the harms that comes from it is not that it falls but that it’s there, ever present over you and killing all of your activities. So I think the Kellner court probably missed the boat. It was a well-intentioned but reckless effort to instill some sort of discipline into the analysis.


Wesley Hodges:  It looks like we do have one question out of the gates. First caller, you are up.


Paul Larkin:  Hi. This is Paul Larkin at the Heritage Foundation. How would you guys treat statements by defendants who just got convicted who then turn to the prosecutor and say something along the lines of either, “When I get out, you’re a dead man,” -- or let’s assume this is a person who is in a managerial position  -- just says “You’re a dead man”?


John Elwood:  It’s seems to me -- I don’t know what Michael’s take is, but it seems pretty open and shut to me because it’s not the case that you have to have the mechanism to take care of it. That is, even if you couldn’t follow through with it, that doesn’t really matter. It seems like a serious expression of an intent to do harm. So to me, it’s not even close. Sure. There’s going to be the fact that time will elapse, but it also, I think, significantly -- it does implicate the purposes for which we penalize threats, which is it engenders fear and it discourages people from doing their jobs. If they acquitted the guy or if they sentenced him light, he would be less likely to make the threat.


Michael Dreeben:  Agree 100 percent.


Wesley Hodges:  Wonderful. Well, Paul, thank you so much for your question. I see no immediate questions. I’ll turn the mic back to Professor Volokh. Do you have any additional thoughts you’d like to move through, or should we go to closing remarks?


Prof. Eugene Volokh:  Let’s go to closing remarks.


Wesley Hodges:  Wonderful. Well, Michael, how about you start us?


Michael Dreeben:  I would like to address very briefly the third question that Professor Volokh put out, which is how to distinguish a threat from a warning? And my idea on this one may not be right but let me put it out so we can have some discussion. If a member of Al Qaeda issues a threat to destroy U.S. troops if they don’t get out of Afghanistan immediately and the New York Times publishes that, has the New York Times engaged in a threat? Or in a more domestic local context, somebody relays something that they overheard somebody else say. I heard John say—not John Elwood—but I heard John say that he’s going to take out the teacher because of a bad grade -- and reports that to the school principal.


I had interpreted the idea behind threats, even though it’s not usually expressed this way in a jury instruction, I think it’s implicit in the jury instruction that it has to be a serious intent to bring about the harm by the speaker, either directly or through some agent who acts at the speaker’s behest. So when the New York Times reprints a threat, everybody knows that the New York Times is not intending to utter words that constitute its own expression of an intent to do anything, even objectively. Whereas the person who’s speaking and actually uttered the words in the first instance is basically talking about drawing on their own power to bring about the harm.


So if you take that as part of the definition of what a threat really is and you look at the speaker’s words, you have to then discount a warning from the threat category because the person who’s issuing the warning isn’t communicating that they will bring about the harm. So that’s what I thought about the third question.


John Elwood:  I think it becomes tricky when the speaker is vague about who is going to do the harm. The example that Eugene gave is that, if police are abusive, someone’s going to take a shot. And that’s vague, and it might be somebody else. It could be a warning, or it could be a threat that that someone might be the speaker. When the act or the remark potentially includes the speaker, that’s I think where it becomes very tricky. And you have to start looking at other indications to get a sense of what exactly the person is attempting to convey.


Michael Dreeben:  There are also three other doctrines that I think are worth talking about. Threats doesn’t have to do all of the work. There’s incitement. There’s solicitation of a violent crime, and there’s conspiracy. And all three of those doctrines have come up in the context of people who have done the kind of thing that Eugene has talked about -- say, if X happens, someone is going to take them out. There was a very well-known notorious case about the so called Nuremberg files, which were put up on a website by opponents of abortion who identified doctors who were performing abortions, several of whom had been killed. And then they would get a red X across them on the website.


And the questions is were they liable for it, or were they simply expressing endorsement or approval or a political view or what? If in fact they have some connection to the people who are actually committing the violent crimes, there may be conspiracy. If they know who they’re targeting, as people who might bring them about, there might be solicitation. And if the immediacy requirements are met, it might actually be incitement. So threats doesn’t have to do all of the work in this area.


John Elwood:  I don’t really have any prepared closing remarks to make. But the one thing that I think is worth noting, in particular, is that as more and more of our lives take place and our interactions with others take place not person to person or face to face but online, it just really increases the opportunities for misunderstanding. People in a conversation in the same room talking face to face, it’s easy enough to have a misunderstanding there. And when you remove so much of the context of the statement, such as tone of voice and expression on your face and various things like that, the opportunity for things to go very badly just increases.


Prof. Eugene Volokh:  Thank you so much to the panelists and to The Federalist Society for hosting this.


Wesley Hodges:  Well, big thank you to you as well, Professor Volokh. On behalf of The Federalist Society, I’d like to thank each of our experts for the benefit of their valuable time and expertise. We welcome all listener feedback by email at Thank you all for joining us for the call. We are now adjourned.


Operator:  Thank you for listening. We hope you enjoyed this practice group podcast. For materials related to this podcast and other Federalist Society multimedia, please visit The Federalist Society's website at