Jennifer E. Rothman
Consider the following scenarios:
You are a physician at a local Planned Parenthood clinic. As part of your job you perform abortions. There have been protests outside the clinic and you have heard about the murders of several doctors around the country who were killed because they performed abortions. One day a colleague calls you and tells you that an anti-abortion group has put up a website which lists the names and home addresses of doctors who perform abortions. When you look at the website you find your name and address on the list along with strong language saying that you and the others on the list will one day be held accountable for your crimes against humanity. Some of the doctors' names have black lines through them. You recognize these names as people who have been murdered by anti-abortion fanatics. Can you successfully sue the creators of the website for threatening you and causing you severe emotional distress, or is this website protected by the First Amendment?
Now imagine yourself a woman in college. You hear from a friend that a classmate has posted a story about you on the Internet with a newsgroup called "sex stories." You read the posting and find a gruesome and detailed story of the narrator torturing and raping you. The story culminates in a description of you being doused with kerosene and lit on fire. The posting uses your real name. You are scared and call the police. Should your classmate be convicted of threatening you?
You attend a rally in support of a boycott of white owned stores whose owners will not hire African American employees. You are aware of several violent acts against blacks who have ignored the boycott including the firing of shots into the house of one boycott violator. The leader of the boycott speaks at the rally and warns boycott violators that "their necks will be broken." You had been considering returning to some of the white owned stores but are frightened by the leader's words. Should the leader of the boycott be arrested for threatening boycott violators or is his speech protected by the First Amendment?
As a child you grew up watching the Lone Ranger on television. From this show you picked up the phrase "the silver bullets are coming" which signified to you that the Lone Ranger was on his way to save the day. Many years later, after an acrimonious divorce, you contact an FBI agent with newfound evidence that implicates your ex-father-in-law in an illegal bankruptcy scheme. On your voice-mail message to the FBI agent, rather than just saying you found new evidence, you use your favorite childhood phrase: "the silver bullets are coming!" Shortly after leaving this message, you are arrested for threatening a federal officer. Should you be convicted?
As the above situations show, there are many different contexts in which statements might be considered threatening. Many courts and scholars have focused only on one or two situations individually. The problem with not considering a broad spectrum of scenarios is that too often scholars and courts rely on gut judgments rather than on a clear and predictable test. The main purpose of this article is to create a test for determining when a statement is a "true threat" not deserving of First Amendment protection.
The law surrounding threats has gained recent attention from commentators after decades of virtual anonymity and unaddressed confusion among the lower courts. The sudden interest in threats has been sparked primarily by the proliferation of widely disseminated Internet speech. In particular, two high profile cases have shined the spotlight on threats: the so-called Nuremberg File case and the Jake Baker case, both of which I used in the above hypotheticals. Despite this recent interest, the three major hornbooks and treatises on the First Amendment and the Constitution still do not have an index listing for true threats.
Various federal statutes make it a crime to convey threats by mail or through interstate commerce, to intimidate or threaten anyone involved in reproductive services, and to threaten the president and other government officials as well as law enforcement officers. Many state statutes also prohibit the making of threats. Outside the arena of criminal law, threats are punished primarily as torts under theories of intentional or reckless infliction of emotional distress.
Defendants in both criminal and civil cases can use the First Amendment as a defense, arguing that their speech should be protected. However, the free speech clause of the Constitution has never been read to protect all speech. Speech such as obscenity, fighting words, child pornography, incitement, and "true threats" is considered outside the protections of the First Amendment. Therefore, the defendant cannot use the First Amendment as a defense for statements which are deemed true threats.
Even though the Supreme Court has made clear that true threats are punishable, it has not clearly defined what speech constitutes a true threat. The only Supreme Court case to elaborate a holding on the basis of the true threats exception to the First Amendment is United States v. Watts, a per curiam decision which made clear that a law "which makes criminal a form of pure speech, must be interpreted with the commands of the First Amendment clearly in mind. What is a threat must be distinguished from what is constitutionally protected speech." However, the Supreme Court did not provide a specific test for making this distinction.
To determine when speech is protected by the First Amendment, and therefore not punishable as a threat, most circuits have adopted either a reasonable speaker or a reasonable listener test. Both these tests essentially boil down to an evaluation of whether or not a reasonable recipient of the statement would believe it constituted a true threat. The Supreme Court has never reviewed the differing circuit court tests to determine their constitutionality or the validity of the circuits' interpretations of Watts.
Most circuits have allowed for the admission of the alleged victim's reaction as evidence of how a reasonable person would interpret the statement. Combined with the reasonable speaker/listener test this makes it possible for people who did not purposely, knowingly, or even recklessly make a threat to be punished for making one. For example, even where the speaker had no expectation that the alleged victim would hear the statement, the speaker can be held liable or convicted of making a threat in most courts.
The reasonable speaker/listener test also allows for the punishment of speech where the alleged threat was simply warning the recipient of a danger. Such warnings, as I will discuss in more detail later, should be protected under the First Amendment not only because they have inherent value when offered with good intentions to warn an individual of danger, but also because such warnings are necessary for the free expression of ideas.
The failing of the reasonable speaker/listener test to allow for warning threats highlights the test's failure to explain the outcome in NAACP v. Claiborne Hardware, the other major Supreme Court case besides Watts directly to address true threats. As I will discuss in detail, the application of the reasonable speaker/listener test would likely lead to a conviction of the defendant in Claiborne Hardware, when in fact the Supreme Court held that the defendant's speech was protected by the First Amendment.
The current confusion among the lower courts and the insufficient protection for speakers can be remedied by the courts. The United States Supreme Court could grant certiorari in a threats case and decisively settle the scope of the true threats exception to the First Amendment. State Supreme Courts are also free to establish clear standards for determining threats under state constitutions which may be more protective of free speech than the United States Constitution. Even federal circuit courts have the opportunity to reexamine their own tests and reformulate them. Thus, it is crucial to consider how the current tests can be reformed to protect both the rights of speakers and of potential victims.
This article proposes a new test for determining what is a true threat. As I will discuss in depth, the reasonable speaker/listener test, adopted by a majority of circuits, is useful but incomplete. Therefore, I add two additional elements to my test: (1) a subjective intent prong which requires the prosecution or plaintiff to prove that the speaker purposely, knowingly, or recklessly intimidated, frightened, or coerced the target; and (2) an actor prong which requires proof that the speaker explicitly or implicitly suggest that he or his co-conspirators will be the ones to carry out the threat. In addition, I develop in more detail the factors that a fact-finder should consider when applying the reasonable listener prong.
The addition of the actor prong is wholly novel and has not been discussed by courts or scholars to date. This prong is crucial to my test, and crucial to the protection of speech under any test for determining whether a true threat has been made. By requiring that there be, at the very least, some implication that the speaker or his associates will be the ones to carry out the threat, greater latitude is given to speakers to use, without fear of punishment, the strong language that the First Amendment allows.
Use of the actor prong will allow speakers to make "warning threats." As I will discuss at more length later in the article, a warning threat is a threat that is offered either for the good of the listener to protect him from harm, or to scare the listener by accurately relaying a danger that exists, but which the speaker has no control or influence over. One of the main ways of distinguishing warning threats from true threats is that in the former, the speaker is not threatening that he or his associates will take any action. In these instances the speech should be protected.
Part I of this article examines the policy behind the First Amendment exception for true threats. Part II traces the Supreme Court's treatment of threats. Part III explores the federal circuit courts' differing opinions of what the Supreme Court intended. Part IV criticizes the current tests applied by the lower courts, as well as some of the alternatives suggested by other commentators. Finally, Part V presents my proposed test and applies it to both hypothetical and actual cases in what I call a "test suite." By applying the proposed test to a variety of situations, I avoid the pitfalls of other tests which only work in a few instances or which have been molded by scholars to reach the desired result in a particular case.
In order to more easily discuss the justifications for making a true threats exception to the First Amendment, a rudimentary definition of speech that could be considered a threat is necessary. At a basic level, true threats are speech which communicates the possibility of future use of physical force or violence against the intended victim or those close to the victim, or unlawful damage of valuable property. I will discuss when such a threat is unprotected by the First Amendment later, but for the moment this definition will provide the basic understanding of what a true threat is.
There are four main reasons why true threats are not protected speech: (1) to protect people from the fear of violence; (2) to prevent the disruption that this fear engenders; (3) to incarcerate people who have identified themselves as likely to carry out a threatened crime before they have the opportunity to perpetrate the crime; and (4) to prevent people from being coerced into acting against their will.
The psychological fear created by a threat to oneself or one's family or the threat of serious property damage (such as the threat to burn down a person's home) is unquestionably a disturbing experience. People who are forced to live under the shadow of such threats suffer a myriad of psychological and health problems including nightmares, heart problems, inability to work, loss of appetite, and insomnia. These very real emotional and physical effects are unjustifiable even by free speech concerns and demand that threatening speech be limited.
Threats can cause major disruptions not only in a victim's life, but also by requiring either public or private resources to be spent on protecting the target from the threatened action. One of the main reasons that threats against the President are punished, even when the likelihood of the threat being carried out is remote, is that the Secret Service must take each threat seriously and spend many hours investigating each person who makes a threat. Justice Marshall elaborated on this danger in his concurring opinion in Rogers v. United States:
Plainly, threats may be costly and dangerous to society in a variety of ways, even when their authors have no intention whatever of carrying them out. Like a threat to blow up a building, a serious threat on the President's life is enormously disruptive and involves substantial costs to the government. A threat made with no present intention of carrying it out may still restrict the President's movements and require a reaction from those charged with protecting the President.
One does not have to be the leader of the country to experience such disruptions. Many people have spent a fortune paying for private security to protect themselves. One example is the abortion providers around the country who are forced to wear bullet-proof vests and hire bodyguards for both themselves and their families. The disruptions are not just financial, but also include changes in behavior, such as not being able to leave one's house or going into hiding to avoid the threatened danger.
The third justification for restricting speech is crime prevention. By punishing a threat we can incarcerate a criminal before he has the opportunity to act. The expression of the threat does not make the threatened action necessarily more likely to be carried out, but it helps law enforcement by identifying the likely perpetrator of a crime. Given this fact, there is no reason to make law enforcement wait until someone is hurt before acting. No one would dispute that the Secret Service should be able to act before someone fires a shot at the President. Even when other people are threatened, there should be a mechanism for protecting them prior to the commission or attempt of a violent act.
Finally, threatening speech may cause the target of that speech to be coerced into acting against his will. In general, people should have the liberty to make choices of their own free will. A threat of violence or of substantial property damage can force people to act differently from how they would want to act. For example, imagine you are a gynecologist who among many other services performs abortions. You are used to protesters and picketers, but when you receive a call from an anti-abortion fanatic who states that he knows where your daughter goes to school and will kill her if you do not stop providing abortions, you are much more likely to stop giving abortions than you would be if no threats had been made. Statutes which punish threats protect people from just this sort of coercion. A statement which causes someone to act against her free will by threatening illegal action should not be protected by the First Amendment.
However, any test for determining which statements are true threats must recognize that some coercive speech is legal and often valuable even if it produces fear or is intimidating. For example, boycotts, strikes, and threats of lawsuits are all legal forms of coercion. As the Supreme Court in NAACP v. Claiborne Hardware emphasizes:
[s]peech does not lose its protected character . . . simply . . . because it may embarrass others or coerce them into action. . . . The claim that the expressions were intended to exercise a coercive impact on respondent does not remove them from the reach of the First Amendment. Petitioners plainly intended to influence respondent's conduct by their activities; this is not fundamentally different from the function of a newspaper.
There is little question that threats of lawsuits and threats to reveal extramarital affairs where no money or other compensation is being extorted are allowable threats. This makes sense not only practically, but also as a matter of policy. If we punish speech merely because it is coercive we will stifle the "uninhibited" and "robust" discussion that the First Amendment seeks to protect. For example, Martin Luther King Jr. threatened in a 1963 letter from the Birmingham Jail that if blacks were not allowed to protest nonviolently using boycotts and marches, violence would break out. This sort of rhetoric helps to foment important social change and should be tolerated despite the suggestion of threatened future violence. This is in accordance with statements from the Supreme Court that some speech which is on its face threatening must be tolerated. Persuasive oratory often becomes heated and is often purposely directed towards coercing, intimidating, and even frightening speakers.
It is necessary to develop a test which distinguishes these types of threats from those which are more harmful than they are valuable. The line is crossed between acceptable and unacceptable coercion when the speaker intentionally or at least recklessly threatens that he or his associates will commit an act of violence or serious property damage against the victim or someone close to the victim.
As the foregoing discussion shows, some threats must be punished. The challenge is to distinguish a true threat from an idle threat, political hyperbole, a jest, misconstrued speech, allowable coercion, or legitimate political advocacy. The following section will explore the small amount of guidance that the Supreme Court has given lower courts on making such a determination.
The only Supreme Court case to provide specific guidance for evaluating true threats is Watts v. United States, a per curiam opinion from 1969. In Watts, the defendant, an eighteen year old man, spoke out during a public rally against police brutality at the Washington Monument in 1966. The defendant said the following to the crowd:
They always holler at us to get an education. And now I have already received my draft classification . . . and I have got to report for my physical this Monday coming. I am not going. If they ever make me carry a rifle the first man I want to get in my sights is L.B.J. . . .
In response to this statement the crowd laughed.
The defendant was convicted of violating 18 U.S.C. § 871, which makes it a crime to knowingly and willfully make any threat to take the life of or to inflict bodily harm upon the President of the United States. The Supreme Court reversed, and held that the defendant had not made a true threat because his statement was political hyperbole.
Given the political backdrop of the rally, the Court recalled the words from New York Times Co. v. Sullivan: "debate on public issues should be uninhibited, robust, and wide-open, and . . . it may well include vehement, caustic, and sometimes unpleasantly sharp attacks on government and public officials." The Court believed that the defendant's statements, rather than constituting a threat, were merely a "crude" way of expressing "political opposition to the President." Even though the Court did not elaborate a test for determining when a threat was protected or unprotected speech, the Court made clear that Watts' statements -- given the context, the conditional nature of the statement (a condition that the speaker had vowed would never come true), and the reaction of the listeners -- were not a true threat.
Thus, the Court in Watts suggested several factors which a court should consider in determining whether or not a statement is a true threat: (1) whether or not the speech constitutes political hyperbole; (2) the overall context in which the statement is made; (3) the reaction of the listeners; and (4) whether or not the statement was conditional, especially if it was conditional on an event that was unlikely to occur. No other Supreme Court case has elaborated on or even applied the factors from Watts.
The only other Supreme Court case that confronted true threats is NAACP v. Claiborne Hardware. Unlike Watts, Claiborne gave no direct guidance on what constitutes a true threat beyond holding that the defendant's speech was not a true threat. The events of Claiborne Hardware took place in Claiborne County, Mississippi during a boycott by black citizens of white merchants. A group of men called alternately the Black Hats, Deacons, or Enforcers stood watch over the stores and took down the names of African Americans who violated the boycott. These names were then read aloud at meetings and published in a newspaper. There were approximately ten violent acts associated with the enforcement of the boycott ranging from the simple trampling of a flower bed to the spraying of bullets into the house of one boycott violator.
There were several issues at stake on appeal, but the one relevant to this discussion is whether or not Charles Evers' public speeches constituted threats. Evers was the field secretary for the NAACP in Mississippi. In a speech on April 1, 1966, before a group of black supporters, Evers allegedly told those assembled that any "'Uncle Toms' who broke the boycott would 'have their necks broken' by their own people. Evers' remarks were directed to all 8,000-plus black residents of Claiborne County . . .."
In a similar speech given on April 19, 1969 before an assembled group at a church Evers stated:
[Y]ou better not be caught on these streets shopping in these stores until these demands are met. . . . Remember you voted this. We intend to enforce it. You needn't go calling the chief of police, he can't help you none. You needn't go calling the sheriff, he can't help you none (That's right.) He ain't going to offer to sleep with none of us men, I can tell you that. (applause).
On April 21, 1969 Evers gave another speech to several hundred people and stated something along the lines of "If we catch any of you going in any of them racist stores, we're gonna break your damn neck."
The Supreme Court of Mississippi found Evers' statements to constitute true threats. The United States Supreme Court reversed. The Court, in an opinion by Justice Stevens, found that Evers' comments were protected speech. The opinion emphasized that just because "expressions were intended to exercise a coercive impact on respondent [that] does not remove them from the reach of the First Amendment." At the same time, the Court made clear that violence and threats of violence are not shielded by the First Amendment.
The Court suggested that Evers' speeches could be considered either as threats or as speech tending to incite unlawful action. The Supreme Court decided without much discussion that Evers' speeches could not be considered true threats and dropped a footnote to Watts. On its face Evers' speech seems like the very "threats of violence" that the Court states are unprotected. However, given that this was a unanimous decision of the Supreme Court, it is important to examine what the Justices uniformly agreed was nonthreatening about Evers' statements.
The context of the speech suggests that the words were meant more as rhetorical hyperbole than as direct, serious threats. The Court pointed to the fact that in each instance he was making a "public address -- which predominantly contained highly charged political rhetoric lying at the core of the First Amendment." Although the Court never explicitly says as much, it seems that the speech could be viewed as the very sort of political hyperbole that the Court in Watts found protected.
The Supreme Court tends to narrowly construe restrictions to the First Amendment in general, but especially in the context of picketing and boycotts where such speech is often viewed as political speech. Even though political speech is itself not protected as a special or separate class of speech, the political backdrop of a speech suggests that the alleged threats are more likely to be rhetorical devices for persuasion rather than serious efforts to threaten listeners with violence. Evers' alleged threats were issued in the context of much longer speeches rallying his supporters behind the civil rights boycott. In fact, two days before the speech in which Evers allegedly said "we're gonna break your damn neck," Evers gave a recorded speech in which he counseled blacks to use nonviolent, political action.
Another factor in the Supreme Court's decision may have been the fact that Evers' audience was not the primary target of the threat. The record is not explicit, but it seems that the majority of the attendees were supportive of the boycott. The audience's approval of Evers' speeches suggests that the people did not view it as a threat to themselves, much as the audience in Watts laughed when the defendant suggested that he would shoot L.B.J. There is no evidence that anyone in the audience was threatened by Evers' speech. Mr. Cutler, the attorney for Evers and the NAACP, pointed out that:
22 black witnesses were called by the merchants and asked about whether they had ever heard of the Evers remark about breaking necks. Sixteen of them said that they had never heard of it at all, and six said they had heard of it only in 1969, three years after the boycott began, and none of the many black witnesses called by these merchants testified as to any fear of physical violence because of the Evers speeches.
The Respondents have only cited to you four instances in which anyone testified about fear of punishment or discipline, and the context of at least two of those statements show that they were speaking of fear of denunciation and ostracism.
Admittedly, it is likely that some potential boycott violators were in the audience and people who did not attend might have later heard Evers' rhetoric recounted in the small black community of Claiborne. However, Evers did not specifically target individual people or the people who were directly in front of him. Evers was certainly aware of the list of boycott violators, but made no reference to them by name. The Supreme Court's holding in Claiborne Hardware suggests that threats towards groups rather than towards specific individuals are more deserving of protection.
Several clues in the parties' briefs and the Court record suggest an additional explanation for the Court's findings. As Mr. Cutler stated in his oral arguments, it was never clear exactly when Evers made the statement about breaking necks and it was never clear exactly what was said. One witness believed that what Evers actually said was: "If you break the boycott your own people will break your necks." This language suggests more of a warning threat of what others might do than a direct threat from Evers. I will discuss warning threats in more detail in Part IV, but as I stated earlier warning threats should be a protected part of legitimate advocacy and expression.
Evers himself believed that he did not say he personally would break anyone's neck. Instead he believed that he said: "their necks needed breaking. I don't think I said I'd break their neck. I'm quite sure I didn't say that." It is quite possible that the Supreme Court was hesitant to find Evers' words threatening when the record was so full of uncertainty regarding what really happened in Claiborne County and what Evers really said. Given the possible racial bias of Claiborne County officials and the lower courts, the Supreme Court may have been reluctant to uphold a finding against Evers, but was nonetheless stuck with the factual findings of the Mississippi courts.
In addition, there was no credible belief that Evers or his associates would be the ones to take action. There was no evidence that any prior violence was carried out by Evers nor that he had ever directed any one else to commit violent acts. In fact, the Supreme Court held that Evers could not be held liable for the actions of others because they found that he was not involved in the commission or the planning of the violence. The fact that Evers was known for preaching nonviolence supports the proposition that no one would believe that Evers meant that he or his associates would take any action. During his testimony Evers stated that "I've always tried to tell my folks that violence is not the way." Evers did talk about blacks disciplining blacks, but not in a violent manner.
Thus, as I have discussed at length above, the Supreme Court's holding that Evers' statements were not true threats makes sense because: (1) his speech was a political speech given in the context of boycotts before a large crowd and most likely was given for rhetorical effect; (2) the audience of the speech was not the primary target of the alleged threat and there was little or no evidence that anyone was in fact threatened by the speeches; (3) the Court could not be certain what Evers said in the 1969 speech which contained the only statement which was on its face threatening; and (4) there was no suggestion either implicitly or explicitly that Evers or his associates suggested that they would take violent action against boycott violators -- in fact, the bulk of Evers' 1969 speech related to his support for nonviolent action.
The Court analyzed in more depth whether or not Evers' speeches met the requirements of the Brandenburg test for incitement. Under this test, the Court found that Evers had not incited imminent lawlessness since no acts of violence followed his speech. The Court emphasized that "[a]n advocate must be free to stimulate his audience with spontaneous and emotional appeals . . . . When such appeals do not incite lawless action, they must be regarded as protected speech." The Court therefore found that without further evidence there was no proof that "Evers authorized, ratified, or directly threatened acts of violence."
Claiborne Hardware, like Watts, leaves lower courts with little guidance because no clear test to determine a true threat is presented. The Court has not granted certiorari in a case on threats since then, leaving the lower courts in the dark as to what test is appropriate.
The Supreme Court's minimal guidance has left each circuit to fashion its own test. The majority of circuits have developed a version of a reasonable person test, but are split over whether the test should be from the perspective of the speaker or the listener. The Second Circuit has split from the pack and adopted a test that adds a requirement of imminence. Some judges on the Ninth and Fourth Circuits think that courts and juries should, in certain circumstances, consider the speaker's intent.
One could argue that allowing each circuit to establish its own test is a good idea. However, there is no justification for regional variations on what speech is punished as a threat. Unlike obscenity law, this is not an area where community standards are particularly relevant. In addition, inconsistent and conflicting standards will chill more speech than would a single, clear, and predictable national standard.
Most circuits employ either a reasonable speaker or reasonable listener test for determining when a true threat has been made. One example of the speaker centered test is set forth in the First Circuit's United States v. Fulmer: "The appropriate standard under which a defendant may be convicted for making a threat is whether he should have reasonably foreseen that the statement he uttered would be taken as a threat by those to whom it is made."
The court in Fulmer also describes why it distinguishes between a reasonable speaker and a reasonable listener test:
This standard not only takes into account the factual context in which the statement was made, but also better avoids the perils that inhere in the "reasonable-recipient standard," namely that the jury will consider the unique sensitivity of the recipient. We find it particularly untenable that, were we to apply a standard guided from the perspective of the recipient, a defendant may be convicted for making an ambiguous statement that the recipient may find threatening because of events not within the knowledge of the defendant. Therefore, we follow the approach of several circuits by holding that the appropriate focus is on what the defendant reasonably should have foreseen.
Even though the Fulmer court makes much of the difference between a reasonable speaker and a reasonable listener test, I believe the distinction is specious. Both tests are essentially a reasonable listener test, because that is how the jury will decide whether a threat was made, regardless of whether they put themselves in the shoes of the speaker or the listener. The only way a jury can evaluate whether the speaker would "have reasonably foreseen that [his statement] would be taken as a threat" by the listener is by considering how a listener would view that statement. To foresee how a threat would be taken by the listener, the only frame of reference a reasonable speaker would have is how the speaker would react if he had heard the statement directed towards himself; i.e. if he were himself a listener. Thus, a reasonable speaker's assessment will boil down to how a reasonable listener would react.
Furthermore, many courts, such as the one in Fulmer, admit testimony about the reaction of the recipient, which makes the reasonable speaker test even more like the reasonable listener test because the jury will rely on how the specific recipient reacted. The reasonable listener test is not more attuned to the eccentricities of the particular recipient of the threat because the reaction must be a reasonable one. Therefore, in an ideal application of the reasonable listener test a particularly sensitive listener's reaction would not be considered "reasonable." Because there is no practical difference between the reasonable speaker test and the reasonable listener test I will consider the circuits which use either one of these tests together.
The Third Circuit uses a reasonable speaker test and like all of the circuits has no requirement that the person intended to carry out the threat. The Fourth Circuit uses a reasonable recipient test in which a jury decides if "an ordinary reasonable recipient who is familiar with the context . . . would interpret [the statement] as a threat." The Fifth Circuit has also adopted a reasonable recipient test. The Sixth and Seventh Circuits have adopted the reasonable speaker test.
The Eighth Circuit focuses on the listener and "whether an objectively reasonable recipient would view the message as a threat." The Ninth Circuit uses the reasonable speaker test. The Tenth Circuit has not outlined as elaborate a test as some of the other circuits, but has agreed that the determination of whether a true threat has been made lies with whether "those who hear or read the threat reasonably consider that an actual threat has been made." Although the Eleventh Circuit has not heard many cases on threats, it also has adopted a reasonable recipient test. The D.C. Circuit has heard only a few cases on threats after being reversed by the Supreme court in Watts, and in those cases, it appears to have adopted the reasonable recipient test.
The Second Circuit in United States v. Kelner interpreted Watts to mean that "[s]o long as the threat on its face and in the circumstances in which it is made is so unequivocal, unconditional, immediate and specific as to the person threatened, as to convey a gravity of purpose and imminent prospect of execution, the statute may properly be applied." The imminence requirement dramatically raises the threshold before speech is considered threatening. This requirement is also combined with a reasonable listener test.
In other words, Kelner requires two elements to be met: (1) the satisfaction of the reasonable listener test, and (2) that the listener believed that threat would be carried out imminently. In addition, like other circuits, the Second Circuit makes clear that it does not matter whether or not the speaker intended to threaten the listener.
The defendant in Kelner was convicted of violating 18 U.S.C. § 875 (c) for transmitting an interstate threat to injure someone. On November 11, 1974, Yasser Arafat, the leader of the Palestine Liberation Organization, planned to be in New York attending a session at the United Nations. The Jewish Defense League organized a demonstration outside the hotel where Arafat and the PLO delegation were scheduled to stay. The defendant, a member of the JDL, was interviewed by a television news crew. The defendant sat behind a desk at the JDL headquarters wearing military fatigues and holding a .38 caliber gun in front of him.
During the interview the defendant said: "We have people who have been trained and who are out now and who intend to make sure that Arafat and his lieutenants do not leave this country alive." When asked specifically if he intended to kill Arafat, the defendant answered "We are planning to assassinate Mr. Arafat." In response to a question about who would carry out the assassination, the defendant answered "[e]verything is planned in detail." This interview aired on the ten o'clock WPIX Channel 11 news.
There was no evidence before the court of whether or not Arafat had any knowledge of the broadcast. At the time of the trial, both the defendant and several witnesses testified that neither he nor any other JDL member had any plans to assassinate Arafat. The defendant claimed that he merely wanted to show the PLO that "we [as Jews] would defend ourselves and protect ourselves."
The Second Circuit rejected the defendant's arguments that his speech was protected by the First Amendment and affirmed his conviction. Specifically, the court rejected the defendant's claim that no communication had been made because the threat had been disseminated to "an indefinite and unknown audience." Instead, the court thought the threat clearly was directed at Arafat. The court also found that there was imminent danger since the defendant said "we have people who have been trained and who are out now . . . ." Despite the defendant's arguments to the contrary, the court did not believe that the defendant's statements constituted political hyperbole.
Even though district courts from other circuits have applied the imminence requirement, no other circuit has explicitly adopted Kelner.
Only the Ninth Circuit and the Fourth Circuit have required a showing that the speaker intended to threaten the target. However, it is not entirely clear that these decisions have been fully adopted by either circuit. Only two cases in the Ninth Circuit have required specific intent by the speaker to threaten. Neither one of them mentioned the First Amendment or Watts. Each based its holding entirely on statutory interpretation of the relevant federal statutes. Even though these holdings are not based explicitly on the First Amendment, the policies behind the courts' interpretations of the statutes are illustrative of doubts among the circuits about the fairness of the reasonable speaker/listener test.
In United States v. Twine, the Ninth Circuit held that there must be a higher threshold for convicting individuals of threatening private citizens or even other public officials than that for the President. Thus, the court in Twine held that §§ 875(c) and 876 require a specific, subjective intent, even though § 875 does not have any mens rea requirement written into the statute.
In United States v. King, the Ninth Circuit reaffirmed the holding in Twine; however, every other true threats case heard by the circuit has relied solely on the reasonable speaker test. Scholars and practitioners for the most part have not even noticed Twine and King as precedents, and it seems that much of the Ninth Circuit also has forgotten about these cases. Thus, it is unclear, at least under these statutes, what the Ninth Circuit's test for a true threat is.
Similarly, the Fourth Circuit in United States v. Patillo required proof that the speaker intended to threaten the victim when there was little likelihood that the victim would receive the alleged threat. Patillo, unlike Twine and King, specifically addresses the holding in Watts and the Supreme Court's guidance on what constitutes true threats under First Amendment analysis. Specifically, the court in Patillo required the defendant to intend to harm or interfere with the President of the United States in order to be convicted when no intent to communicate the threat is present. Patillo is the only case to hold that 18 U.S.C. § 871 and the First Amendment require not only the specific intent of the speaker to threaten, but also the "present intention" to carry out the threat where there is no communication or attempted communication of a threat to the President. However, no other case in the Fourth Circuit has used this requirement. Despite the infrequency of other courts' reliance on Patillo, Twine, and King, neither the Fourth nor the Ninth Circuit has ever explicitly overruled these cases.
The Sixth Circuit in United States v. Alkhabaz (the "Jake Baker" case) held that speech can only be a punishable threat if it is directed at the achievement of a specific goal or coercive impact on the target. Jake Baker was tried for sending threatening communications in interstate commerce, by exchanging e-mails with an Internet pen pal "which expressed a sexual interest in violence against women and girls." Baker had initially also been charged with making a threat by posting a story on a popular newsgroup website called "alt.sex.stories." This story described the brutal torture, rape, and murder of a woman who he gave the same name as one of his University of Michigan classmates. The court did not directly address the posted story because the government did not pursue this claim in the district court.
The Sixth Circuit affirmed the district court's exoneration of Baker The Sixth Circuit argued that "[a]t their core, threats are tools that are employed when one wishes to have some effect, or achieve some goal, through intimidation." Specifically, the court pointed to "extortionate or coercive" threats. Thus, in addition to the reasonable listener test, the Sixth Circuit added a requirement that the threat be perceived as "communicated to effect some change or achieve some goal through intimidation."
It is unclear what sort of goal the Alkhabaz court meant to require. The language of the case suggests that the goal of merely frightening the target is not sufficient to constitute a threat. However, because the communication at issue was not likely to be received, it is possible that the court found that no goal, including one of frightening anyone, could be realized. The holding makes sense to the extent that the court meant to include frightening a recipient as a goal -- although the court could unquestionably have expressed itself more clearly.
However, if frightening a victim is not enough to constitute a goal, the logic of the holding is deeply flawed. Often a threat's sole purpose is to scare the listener, rather than to convince the listener to take certain actions. As I discussed in Part I, fear causes two of the main harms from threats: the psychological and physical effects, and the disruption caused by taking preventative measures. The holding in Alkhabaz has yet to be applied in another case in the Sixth Circuit, so it is not even clear how its holding will be interpreted by the Circuit itself.
The existing lower court tests are vague, at times insufficiently protective of speech, at other times overprotective of speech, and inconsistently applied by judges and juries.
The absence of a requirement that the speaker intend to threaten, or at least act knowingly or recklessly, leads to a high probability that speech will be punished when it should not be. United States v. Fulmer highlights the dangers of using only a reasonable speaker/listener test.
Fulmer was convicted in the District Court of Massachusetts for threatening a federal agent in violation of 18 U.S.C. § 115(a)(1)(B). After an ugly divorce, Fulmer contacted an FBI agent, Egan, regarding alleged pension and income tax fraud by his ex-father-in-law and brother. After some investigation, an Assistant United States Attorney decided not to prosecute the case. Egan told Fulmer of the agency's decision.
There was no further contact until four months later when Fulmer called Egan and left the following voice-mail message:
Hi Dick, Kevan Fulmer. Hope things are well, hope you had an enjoyable Easter and all the other holidays since I've spoken with you last. I want you to look something up. It's known as misprision. Just think of it in terms of misprision of felony. Hope all is well. The silver bullets are coming. I'll talk to you. Enjoy the intriguing unraveling of what I said to you. Talk to you, Dick. It's been a pleasure. Take care.
The FBI agent, Egan, was "shocked" by this message and found it "chilling" and "scary." He viewed the use of the phrase "the silver bullets are coming" as a threat. Various witnesses testified that Egan was upset and Egan stated that he loaded his gun fearing an attack.
Despite Egan's reaction, evidence showed that Fulmer used the phrase "silver bullets" to describe "a clear-cut simple violation of law" -- a phrase similar in usage to the expression the "smoking gun." Two witnesses stated that the defendant used the phrase to describe specific evidence that he had found implicating his father-in-law and brother. In fact, he had discovered an $8,300 check from a bankruptcy estate that had never reached its intended recipient. This check was allegedly the impetus for his phone call.
The First Circuit overturned the conviction based on evidentiary failings and remanded the case, but approved the District Court's use of jury instructions describing the reasonable speaker test. Although the defendant eventually was acquitted, too much is left to chance under the reasonable listener/speaker test because reasonable juries might come to different conclusions about Fulmer's speech under such a test. One could imagine a jury deciding to convict based on the FBI agent's subjective testimony and their own subjective interpretation of how a reasonable person would have understood the phrase "the silver bullets are coming." As the initial jury decision in Fulmer suggests, there is a danger that ambiguous statements not intended to be threats will be interpreted as threats under the reasonable speaker/listener test.
One of the main failures of having only a reasonable speaker/listener test is that it does not incorporate any consideration of the speaker's actual intent. Justice Marshall highlighted these dangers in his concurring opinion in Rogers v. United States, where he criticized the circuit courts' adoption of a construction of 18 U.S.C. § 871 and Watts that allows the conviction of anyone whose statement could reasonably be understood as a threat against the President.
Even though Marshall's concurring opinion was based primarily on his view of the statutory interpretation of § 871, the policy behind his views applies to any punishment of speech as a threat. Marshall made clear that the dangers of encroachment on the First Amendment are not limited to the political arena: "Although the petitioner in the present case was not at a political rally or engaged in formal political discussion, the same concern counsels against permitting the statute such a broad construction . . . . [T]here is a substantial risk of conviction for a merely crude or careless expression of political enmity."
Marshall points to several harms that come from relying solely on a reasonable listener test. He warns that the reasonable speaker/listener test or what he calls the "objective" standard punishes people criminally under a negligence standard -- something that the Court "has long been reluctant" to do -- and emphasizes that the Court should be especially reluctant to do so given the fact that the statute regulates pure speech. A negligence standard is insufficient to protect speech because a speaker may have difficulty telling in advance what will be construed as a threat by a jury, and may be deterred from speaking even where his speech is not negligent. Furthermore, a speaker who thought that what he was saying could not be viewed as threat, but was mistaken, could be punished.
Punishing merely negligent speech will chill legitimate speech by forcing speakers to steer clear of any questionable speech. This would discourage "the 'uninhibited, robust and wide-open' debate that the First Amendment is intended to protect." Although some people might be deterred from speaking freely by a standard of recklessness or more, there will be a less severe chill on speech than under a negligence standard. Furthermore, the goal of deterring threatening speech will not be achieved if a speaker is unaware, in other words negligent, that what he said could be viewed as a threat.
A negligence standard is not enough to protect a speaker's rights. Individuals who poorly communicate or who unknowingly use words which could be misconstrued should not be punished for their lack of oratorical skills when they do not speak with the purpose, knowledge, or with reckless disregard of making a threat. To punish the speech without proving the speaker's intent would create a great danger of discouraging speakers from expressing their views even when such views are on important matters of social and political concern. Some may argue that a negligence standard is appropriate because it forces people to think before speaking. However, given the vagueness of the current test and the variations in determinations by juries, a speaker may not be able to accurately predict the jury's determination of what is reasonable.
Furthermore, in most areas of First Amendment law mere negligence is not enough to punish pure speech. For example, the Brandenburg test for incitement requires that the speaker intend or "direct" his efforts at inciting imminent lawlessness. The Court in New York Times v. Sullivan held that in order for a public figure to recover for defamatory falsehood, he must show that the defendant spoke with "actual malice" which requires a showing that the defendant either knew that the statement was false or spoke with a reckless disregard for its truth. Even when the false statement is spoken about a private figure on a matter of public concern, "actual malice" is required to collect punitive and presumed damages. This higher standard is in place because punitive damages are more akin to criminal punishment and should not be awarded for mere negligence on the part of the speaker.
Justice Marshall is not the only jurist to have supported the addition of an intent requirement. As discussed earlier, at least a few judges in the Fourth and Ninth Circuit believe that juries should consider the speaker's subjective intent to threaten under certain statutes and in certain circumstances. Judge Logan's dissent in United States v. Crews suggests that where speech is unlikely to be received by the alleged victim, the courts should consider the addition of a subjective intent requirement. Several legal scholars and practitioners also have suggested the addition of subjective intent to the test for true threats. 
One of the main reasons an intent standard has not been adopted by the courts is the fear that a requirement of proving the speaker's subjective intent will make it harder to convict. However, most crimes have a mens rea element which requires that the prosecution prove the defendant's state of mind. No one has argued that the requirement of proving a defendant's state of mind should be eliminated from other crimes because it is too difficult to prove. As discussed above, First Amendment law often requires proof of a specific state of mind before allowing the conviction or finding of liability of a speaker.
Almost every Circuit that uses the reasonable speaker/listener test allows in evidence of the target's reaction to demonstrate the likely reaction of a reasonable listener. In Fulmer the court upheld the use of reaction testimony, explaining that "[t]he actual recipient's reaction to the statement shows that the recipient did perceive the message as a threat. This reaction is probative of whether one who makes such a statement might reasonably foresee that such a statement would be taken as a threat."
The admission of such reaction testimony undercuts the supposed objectivity of the reasonable person test by incorporating the reaction of a potentially overly sensitive listener. For example, in Fulmer, the FBI agent's fear that his life was in danger, and his taking of precautionary measures such as loading his gun, partly was responsible for the first jury's conviction of the defendant. It does not make sense to apply tort law's "eggshell skull" plaintiff doctrine in the arena of First Amendment protections, otherwise overly sensitive listeners would be able to suppress all kinds of speech, not just threats.
Under only a reasonable speaker/listener test, the jury is likely to be heavily swayed by the reaction of the recipient of the statement. It may in fact be useful to consider the recipient's reaction as some evidence of how someone close to the situation would have interpreted the speech. This is especially true given the importance of context and tone in threats analysis. However, without the protections that the addition of an intent requirement would afford, the admission of recipient reaction testimony is overly prejudicial because it may have a tendency to be overweighted by juries. This is true because members of a jury will trust the recipient's characterization over their own instinct since the recipient was actually there. In addition, members of the jury are likely to be swayed by the emotional impact of the recipient's testimony.
The test elaborated by the Second Circuit in United States v. Kelner adds an imminence requirement to true threats doctrine. The addition of an imminence requirement overprotects threatening speech by requiring the listener to believe that the threat will be carried out almost immediately. This is antithetical to the whole point of punishing threats. The main concerns, as discussed earlier, are the emotional distress caused by threats, the cost of the precautions they engender, and the coercive effect they may have. Regardless of whether the threat will be carried out imminently, the individual's fear and distress will be the same since the victim will not be able to predict with accuracy when the speaker will strike.
Judge Mulligan points out, in his concurring opinion in Kelner, that even if the threat was to kill Arafat in one week, the speech should still be punished as being a threat. Similarly, if a person told you "I might kill you next week, or maybe next year, but don't worry, I'll get you," this should not be protected regardless of the timeframe.
Even though the Supreme Court has used an imminence requirement to evaluate incitement, the use of such a requirement is not warranted when evaluating threats. The use of an imminence requirement to protect a speaker is justified only when the relationship between the speech and the threat of violence is more tenuous. This makes sense because in these instances courts examine not whether a specific threat was made but rather whether the speech would incite others to act unlawfully. Since the speaker is generally not in control of what others who hear his speech do, and is not threatening action of his own, the speaker deserves the added protection of an imminence requirement. In contrast, when a court evaluates whether a true threat has been made, there is a closer relationship between the speaker's statement and the potential for violence; therefore alleged threats do not merit the additional protection of an imminence requirement.
No court has explicitly considered in any of their tests whether the speaker or his associates will be the ones to carry out the threatened action. This is a crucial element of my test and is the key to developing a clear and uniform test for true threats which is faithful to Supreme Court precedent. If one does not consider who the speaker suggested would take action, three types of valuable speech may be restricted.
First, strong rhetoric which uses such language as "one oughta have one's necks broken" might be wrongly categorized and punished as a threat. The dangers of suppressing such speech are great. As speeches given by civil rights leaders demonstrate, strong rhetoric, even that which on its face seems threatening, is an invaluable tool of advocacy. Martin Luther King Jr. once stated:
[I]f our white brothers dismiss as 'rabble-rousers' and 'outside agitators' those of us who employ nonviolent direct action, and if they refuse to support our nonviolent efforts, millions of Negroes will, out of frustration and despair, seek solace and security in black-nationalist ideologies, a development that would inevitably lead to a frightening racial nightmare. . . . The Negro has many pent-up resentments and latent frustrations, and he must release them. So let him march; let him make prayer pilgrimages to the city hall; let him go on freedom rides . . . If his repressed emotions are not released in nonviolent ways, they will seek expression through violence; this is not a threat but a fact of history."
Most people would agree that this is the very sort of speech that the First Amendment protects. However, on its face it could be construed as a threat that if King and his followers are not supported violence will break out. Such rhetoric is an important part of the dialogue in all civil rights movements. As the Supreme Court in Claiborne Hardware demonstrates, the Court is willing to accept violent rhetoric even against a backdrop of violence, in order to preserve robust and unfettered debate. Speech which contains violent rhetoric is more properly considered under incitement law than as a true threat.
Second, warning threats made with altruistic motives might be restricted unless one considers who would carry out the threatened action. An altruistic warning threat informs the listener or target of a danger without threatening that the speaker or his associates will take action against the target. Such threats serve a useful function because they can warn people of danger. For example, consider a warning given to a woman who is entering a bad neighborhood from a man on the street. He might yell out at her: "I wouldn't go that way if I were you. Someone's liable to kill you." The speaker here is not threatening to kill the woman himself, but rather is trying to help her out by warning that others in the neighborhood might.
Few would debate the proposition that we should allow people to warn others about dangers they might face if they take or fail to take certain actions. Even though this speech might frighten the listener, it is in the listener's interest to be apprised of dangers she may face when those dangers are not a result of actions that the speaker himself will take.
Third, coercive warning threats which point to potential threats from others with the purpose of persuading the targets might be restricted. For example, consider the preceding hypothetical: would it change things if the speaker in the neighborhood was white and the woman was black, and the speaker wanted to discourage black people from coming into his all white neighborhood? Such warnings are similar to the more altruistic ones, in that the speaker is not suggesting that he or his associates will take action, but here the speaker is trying to coerce rather than to aid the target. Such warning threats can constitute permissible political advocacy even if they are not always admirable.
This scenario is more similar to the situation in Claiborne Hardware where Evers did want to intimidate and coerce his audience into obeying the boycott, but the Supreme Court found his actions permissible. Coercion and intimidation are acceptable forms of speech so long as they do not directly threaten their intended targets with violence, destruction of property, or extort money or other compensation. To argue otherwise will result in censoring much civil rights advocacy, including statements such as that of Martin Luther King Jr. discussed above.
Even though the Court, in Claiborne Hardware, was never explicit about why it decided that Evers' speech was not threatening, one of the most convincing explanations is that Evers and his associates were not the ones committing acts of violence and at least two of his alleged threats did not imply that he or his associates would themselves take action. When Evers suggested that the sheriff would not be able to protect people, he was arguably warning potential boycott violators of what some passionate individuals whom he had no control over might do to an individual who broke the boycott. Similarly, Evers' speech in 1966 that boycott violators would "have their necks broken" by their own people is a warning of what others would do to boycott violators with or without Evers' permission. As such, even if it has a coercive or intimidating effect, the speech was not a true threat. These sorts of statements constitute warning threats which should be allowed under the First Amendment because they are a valuable and necessary part of the advocacy of ideas.
The Nuremberg Files can also be seen as a coercive warning threat. In Planned Parenthood of the Columbia/Williamette v. American Coalition of Life Activists ["ACLA"], a federal jury in Oregon awarded Planned Parenthood $107 million in damages because of the ACLA's dissemination of two anti-abortion posters (the so-called Wanted Ads) and its creation of an anti-abortion website now commonly referred to as the Nuremberg Files. For the purposes of my discussion, I will focus on the Internet site and not the posters.
The Internet site listed the names of people who provided or supported abortion services including doctors, nurses, clinic workers, police, politicians, and judges. Along with the names, the website included home addresses and telephone numbers for the abortion providers. Some of this information was publicly available, such as the names of the doctors and phone numbers and addresses listed in the telephone book, but the rest of the information was not readily available.
The list was prefaced with rhetoric stating that the listed people would one day be tried for their "crimes against humanity" much as the Nazi war criminals were tried before the court at Nuremberg. The webmasters also drew lines through the names of abortion providers who had been murdered or wounded. The District Court of Oregon held that this could be found a true threat by the jury and denied a motion for summary judgment on the pleadings.
This might seem like the right conclusion under the reasonable listener test, but it does not square with the directives of the First Amendment. The website creators had a First Amendment right as anti-abortion advocates to publicize the names of abortion providers and to encourage people to picket, boycott, and protest against these individuals.
Does the mere fact that the names were crossed out change the protection of this speech? Was the act of crossing out the names a threat? Given that the names were crossed out after the fact it seems more likely that the crossing out was a recording mechanism of a past event rather than a future threat. The crossing out of names certainly is morbid, but, unless additional evidence is present, it seems to be more in the nature of publicizing and praising the death or wounding of abortion providers rather than any type of prediction of who the next victim will be.
It is true that some of the defendants were avowed supporters of violence against abortion providers. However, there was no evidence that the website creators, who were a subset of the overall defendants in the Planned Parenthood cases, had any history themselves of violent activities. In addition, there was no evidence that the website creators (or anyone associated with them) had any intention of threatening that they or their associates would take action against the abortion providers. Nor were they reckless as to whether a reasonable person would infer that they would personally take action since there was nothing on the website or elsewhere to signify that they would take any action.
Even though it may not be admirable for an anti-abortion activist to be happy that an abortion doctor was murdered, it is certainly his prerogative. In fact this is not very different from the speech protected in Rankin v. McPherson, which upheld a government employee's right to say after a failed presidential assassination, that "if they go for him again, I hope they get him." Flipping the politics of the scenario might change some people's perspective -- suppose that a group of abortion rights advocates create a website which lists the names and addresses of radical anti-abortionists. When some of the radicals are murdered, the webmaster crosses out the names and places a happy face beside it. Again, this may be in poor taste, but should be legal nonetheless.
Even though the makers of the Nuremberg Files may well have intended to intimidate the abortion doctors, there is no evidence that they intended to do so by threatening that they would take illegal action against the listed individuals. The information posted on the Nuremberg Files seems primarily to have facilitated anti-abortion protests by singling out pro-abortion individuals and doctors who should be targeted for protests. Such speech is an important part of the public debate on a controversial issue, and banning such speech under a threat theory would jeopardize substantial political rhetoric which is found valuable in other contexts. Furthermore, the First Amendment clearly protects threats of picketing and boycotts.
Critics of including an actor prong may argue that it is unnecessary to protect warning threats because most warning threats are not prosecuted. This argument places too much faith in the hands of the prosecutors and their discretion. As highlighted by the Nuremberg Files, where heated rhetoric is used in the name of political advocacy, prosecutors will likely feel political pressure to prosecute such warning threats. In addition, prosecutors may use their own viewpoint based preferences to decide which warning threats to prosecute and which to ignore. Therefore, the only way to adequately protect First Amendment speech is by establishing clearer protections for warning threats.
Even though at first glance warning threats might cause some of the same harms that are sought to be protected by banning threats, such as producing fear and forcing individuals and officials to take protective measures, warning threats can be and must be distinguished from true threats. To the extent that warning threats are as coercive as true threats, this is a harm that must be tolerated because the risk of erring in the other direction is too great. Just as we must accept the advocacy of some violence under Brandenburg, we must accept some coercion in the name of keeping open the channels of public debate.
Finally, as seen in the Nuremberg Files and Claiborne Hardware, without the requirement of an actor prong there is a danger that ambiguous threats (those which are not explicit) might swallow up much of political advocacy. If any strong rhetoric spoken against a backdrop of violence can be read as a threat, prosecutors and foes of certain viewpoints will be able to suppress much political and controversial speech. It is this type of fear that led the Supreme Court in Brandenburg to add an imminence requirement in incitement doctrine, because otherwise the advocacy of unlawful action exception to the First Amendment would encompass much of public debate and speech. Thus, for the reasons outlined above, the use of the actor prong protects speakers from overzealous prosecutors, the tides of public opinion, and the misconstruing of ambiguous rhetoric.
If a speaker makes a threat which is not likely to be received by the target, this speech should be protected. The fact that the speaker chooses to convey the threat in a forum in which the intended victim would never hear the threat suggests that the speaker did not purposefully, knowingly, or recklessly attempt to intimidate, coerce, or frighten the alleged target.
In United States v. Crews, after a mental patient watched "The Day After," a disturbing television movie about nuclear annihilation, he said he would kill President Reagan. Even though this was a threat against the President, the speaker made no effort to communicate the threat to the President nor to anyone who he would think was likely to convey the threat. The Court of Appeals vacated and remanded on other grounds, but found nothing about the situation to prevent the defendant from being convicted.
It seems unreasonable to find the defendant guilty in this instance since he was primarily speaking to the television set or possibly a nearby nurse. He had no awareness that his words would be conveyed to the President or even to the authorities. Even though the Secret Service did in fact come out to investigate, it was unlikely that the patient suspected that the psychiatric nurses would call the authorities in this situation. Furthermore, the Secret Service did not expend much time or money by sending one agent to interview the defendant. The dissent by Judge Logan argues that:
. . . When a patient in the psychiatric ward of a hospital utters to a nurse who is treating him a threat against the President, and that patient has been undergoing treatment expressly designed to help him control such outbursts, it is unreasonable to assume that he intended or believed that the threat would be reported to law enforcement authorities.
Also consider the Jake Baker case, where the defendant was charged for threatening speech about third parties sent in private e-mails to a friend. These types of private communication should not be considered threats since it is highly unlikely that they will reach the potentially threatened party. The simple fact that the speaker chose to convey the alleged threat in a way calculated not to reach the threatened party suggests that the speaker did not intend to threaten the target.
Furthermore, three of the main harms sought to be prevented by punishing threats would not occur if the recipient never receives the threat since the recipient would not take precautions, be coerced, or frightened. Even though the justification of preventing crime may be limited by a requirement of likelihood of receipt, the First Amendment case law is clear that the mere expression or even advocacy of violence is not a sufficient basis to punish speech. In other words, just because people may have identified themselves as potential criminal actors by expressing a violent thought or threat, unless there is more evidence against them, they should not be punished for their words. Despite the logic of considering the likelihood of a threat being received, only the Fourth Circuit, as discussed earlier, has held that where there is no likelihood of receipt of the threat the courts should consider the defendant's subjective intent.
There is some confusion among the lower courts about whether incitement law or threats law should be applied to seemingly threatening speech. Even though it is true that courts can apply both doctrines to threatening speech, the two tests are clearly distinct. One scholar, Steven G. Gey, believes that all threats should be analyzed using the Brandenburg test. In fact, Gey interprets the Supreme Court case law, including Watts, to suggest that not only must the speaker intend to act on the threat, but the speaker must also intend to carry out the threat immediately; otherwise, the speech should be protected. This is a controversial interpretation of the case law. The Supreme Court's analysis of threats in NAACP v. Claiborne Hardware suggests that the requirements of Brandenburg are not the test for threats, but that they apply when no true threat has been shown.
In addition, Gey's suggestion fails to take into consideration the policy behind the punishment of threats. Gey's focus on whether or not the speaker actually intends to carry out the threat is irrelevant. He states that "[t]he expression of a desire that a particular person suffer harm or even death is not enough to support legal action against a speaker if there is no evidence that the speaker is taking action to carry out that desire." Gey is in part correct that the government should not be in the business of punishing people for merely wishing someone were dead, however a threat to cause someone harm, regardless of whether the speaker has any intention of carrying out the threat (imminently or not), must be punished since it creates fear, and may lead to coercion and to the spending of resources to prevent the threatened harm.
Some scholars have suggested that different levels of protection should be offered for threats depending on whether or not the threat is uttered in a political context. Even though Watts and Claiborne Hardware suggest that the Supreme Court provides greater protection for political speech, this is somewhat of a red herring. The distinction is not, as many suggest, between political and nonpolitical speech. The use of political speech as a dividing line for any test to determine what is a true threat will unnecessarily complicate rather than simplify threat analysis. First of all, it is difficult to determine exactly what is political speech. It is obvious that a threat from a jilted girlfriend to an ex-boyfriend is not political speech, but it is not as clear whether cross burning is political speech which the speaker could be trying to use to make a statement about race relations.
Obscenity law highlights the difficulty of making such distinctions. The infamous formulation "I'll know it when I see it" is best left out of the law of threats. Justice Brennan explains the failings of obscenity analysis: "Although we have assumed that obscenity does exist and that we 'know it when (we) see it', we are manifestly unable to describe it in advance except by reference to concepts so elusive that they fail to distinguish clearly between protected and unprotected speech." Brennan outlines the dangers of such an uncertain test: some speakers will be discouraged from speaking and othesr will be punished without prior notice that what they were doing was wrong. Similarly, it is difficult to tell what is political and what is not political speech. Like obscenity, individual judges would have to make arbitrary determinations based on what they think is political or nonpolitical. Just as speakers are chilled by obscenity law, speakers will be chilled from making even more valuable expressions for fear that their speech will be categorized by some judges as not being political enough to warrant protection.
Furthermore, speech that is not political still can be valuable in the exchange of ideas, communication, and for self-expression. Other areas of the law such as obscenity allow speech to be protected not only if it has serious political value but also if it has literary, social, or artistic value.  In addition, not all political speech is protected. Consider that many, if not all, threats against the President are political speech but are not generally protected.
Under the proposed test there are no arbitrary determinations of what is political speech and what is not. The same test applies to all scenarios. Even so, the backdrop of a political rally might lend credence for determining, given the context, that words that sound a lot like a threat may not be one. Thus, political speech may be useful shorthand to suggest that in the context a statement was mere rhetoric for persuasive rather than threatening purposes. This is true because the speaker is often speaking to supporters and may be exaggerating for effect as was the case in both Watts and Claiborne Hardware. The litmus test, however, cannot be whether or not a statement is related to a matter of political concern.
Some commentators and judges have suggested that there should be more stringent or different rules for speech on the Internet. For example, the district court in United States v. Baker suggested that the Internet may require new or modified laws. However, there is no reason to treat threats differently depending on the medium in which they are conveyed.
As my discussion in Part V will demonstrate, the same test can be applied for statements made both on and off the Internet. It is true that the Internet can convey information, and hence a threat, widely and quickly, but there is nothing else that makes an on-line threat more threatening. Hence, there is no reason to restrict speech on the Internet any more or less than speech conveyed using other forms of communication.
A host of commentators have suggested different approaches to evaluating threats. Many create unwieldy tests with so many prongs and factors that one cannot imagine a jury sifting through all of their parts. Some suggest specious distinctions. For example, Donald Downs suggests in Nazis in Skokie that threats of violence against people based on race and ethnicity should be per se exceptions to the First Amendment. This is an unnecessary complication of the analysis. An adequate test for determining true threats will protect victims regardless of who those victims are and regardless of the motivations of the speaker.
John Rothchild in Menacing Speech and the First Amendment suggests the creation of a new exception to the First Amendment which would punish "menacing speech," namely speech advocating violence that causes the harms normally associated with true threats but which does not meet the requirements of a true threat. This would allow the government to punish speech which does not qualify as a true threat or meet the Brandenburg requirements for inciting violence, but which Rothchild feels should be punished nonetheless. Rothchild does not articulate a specific test but rather suggests either the weakening of the Brandenburg test or the factors from Watts to include speech which though not directly threatening might cause fear in some listeners. It is only with great trepidation that the courts should read into the First Amendment additional exceptions. The more exceptions that are read into the First Amendment the less power the free speech clause will have. A new exception to the First Amendment should only be created where existing exceptions are wholly inadequate and the harms are very great. Here the exception for threats can take care of menacing speech without limiting more speech than is necessary. The creation of another category for menacing speech is essentially a way for Rothchild to allow punishment of speech he does not like, such as the Nuremberg Files.
Jeremy C. Martin in Deconstructing 'Constructive Threats' argues that the courts should simply do a better job of applying the factors elucidated in Watts. As should be clear from the earlier discussion, one of the main reasons there is so much confusion is that the factors in Watts do not provide enough guidance for courts.
In addition to complicating the analysis of threats by creating more than one test depending on whether the threat is explicit or implied, Martin also complicates the already insufficient reasonable speaker/listener test by restating it as a 3 prong test for implied threats: 1) a reasonable speaker would foresee the expression would cause intimidation or fear; 2) a reasonable listener would be intimidated or afraid; and 3) the target of the threat is actually frightened. As discussed earlier, prongs one and two are the reasonable speaker and the reasonable listener test, which boil down to the same analysis by the jury. The third prong is an addition, but will not change much given that testimony about the recipient's reaction is nearly universally admitted as evidence of how a reasonable person would react.
As these commentators' efforts suggest, not only have the lower courts been unable to develop a workable test for threats, neither have legal scholars.
I propose a three part test for determining whether a "true threat" deserving of punishment has been made. The test elaborates the factors a fact-finder should consider when using the reasonable listener test and adds two new elements: (1) the intent of the speaker, and (2) whether the speaker suggests that he or his associates will be the ones to carry out the threatened action. The test is as follows:
1) Intent prong: The speaker must have purposely, knowingly, or recklessly made a statement which would intimidate, frighten, or coerce the victim(s) with the threat of physical force, violence, or destruction of substantial property. It is irrelevant whether the speaker intended to carry out the threat. Factors that aid in determining intent include:
a. Whether it was likely that the threat would reach the target or his associates. Actual receipt is irrelevant.
b. Whether the speaker told others of his intentions.
c. Whether the speaker made prior threats.
d. Whether the alleged threat violates the speaker's known beliefs, such as an avowed belief in nonviolence.
e. Whether the statement is rhetorical hyperbole. This is often true in political advocacy, where the speaker is exaggerating for effect and there is a suggestion that the speaker's statement was not meant to be taken literally.
2)Actor Prong: The speaker must have purposely, knowingly, or recklessly suggested either explicitly or implicitly that the threat would be carried out by either the speaker or his co-conspirators rather than by unrelated third parties.
3)Reasonable Listener Prong: Considering the entirety of the context, a reasonable person who heard the statement would conclude that the statement was meant to threaten the target or someone close to him with violence or damage to valuable property. The following are factors which are suggestive of such a serious threat:
a. The threat is specific. The less specific the target the less likely it is that the speech is a threat. The more specific the time and method of the action threatened the more likely it is the speech is a threat.
b. The statement is taken seriously by the listeners.
c. The statement is not conditioned on an unlikely or impossible occurrence.
d. The statement is made in an atmosphere of violence.
The best way to evaluate whether the above test is superior to those tests which are currently applied is to apply the test to a suite of cases both real and hypothetical. I have discussed many of the justifications for the proposed test in the prior section, but I will briefly summarize here the basis for each element of the three prong test. There is a need to add an intent prong to protect speakers who are unaware that others might construe their statements as threats. There is also a need to add an actor prong to allow for speakers to use strong rhetoric for advocacy, as well as to allow for warning threats of both varieties -- the coercive and the altruistic. My adaptations to the reasonable listener test simplify and clarify the test by making it easier for jurors to apply. I make explicit what seems implicit in the case law by outlining specific factors which should be considered in applying the reasonable listener test. For example, consideration of the specificity of the threat protects speech which is hyperbolic or used for rhetorical effect.
There is some confusion about whether juries or judges are the ultimate arbiters of the application of the test for threats. As in other areas of First Amendment law, independent appellate review should apply to the analysis of what constitutes a true threat. Even though Bose Corp. v. Consumers Union of United States did not address threats directly, the same policy underlying that decision strongly suggests that independent appellate review should be applied when evaluating true threats. As the Court in Bose emphasizes, it is important that judges have the final word on whether a threat has been made under the proposed test. This insures that the law will be consistently applied, affording the greatest protection for speakers because they will be able to predict what speech is and is not protected.
I briefly address this very large topic only to clarify the procedural considerations involved in applying my test. The test should be used by trial judges in determining summary judgment motions, motions to dismiss, judgments not on the verdict, and by appellate judges when conducting de novo review. The test can also be turned into jury instructions for trial juries to apply, but as with other First Amendment questions, the final word should be that of a judge conducting independent appellate review. I take this perspective as I apply the test to each case in the test suite.
All courts agree that a direct threat, which is not said as rhetorical hyperbole or in jest and is not a highly conditional statement, is unprotected. The outcome would be the same under the proposed test. Take as an example a situation where Kelly, a spurned ex-girlfriend, calls up Randall and tells him that he better watch out because she bought a new rifle and wants to test it out on him. This speech is not protected under current law and is not protected under my proposed test.
The intent prong is met since the explicit nature of Kelly's words makes clear that she purposely threatened Randall. There is no question that Randall not only was likely to receive the threat but also could not avoid receiving it since Kelly directly spoke with him. Even though there was no prior history of threats or violence there would be little doubt that Kelly meant, at the very least, to frighten Randall.
The actor prong is met because Kelly stated that she would personally act. The reasonable listener prong is also met since there is no doubt that a reasonable listener would view Kelly's statement, given the context, as a serious threat. The threat had a specific target, and nothing indicated that it was said in jest.
Thus, Kelly's threat would not be protected by the First Amendment under the proposed test. This result fits with the broad agreement by courts that these types of clear, direct, and personal threats are unprotected. Such direct threats have little or no value and produce all of the harms that threats cause. This example, therefore, is the prototypical example of why there is a true threats exception to the First Amendment.
In United States v. Kelner, the defendant, a member of the Jewish Defense League, threatened to assassinate Yasser Arafat while Arafat was in New York for a meeting at the United Nations. The Second Circuit found the defendant guilty of making a true threat. Under my proposed test the outcome would be the same.
The intent prong is met. Not only did Kelner admit that he wanted to intimidate Arafat, but based on the explicitness of Kelner's words alone a jury or judge is likely to find that he either purposely or knowingly sought to intimidate Arafat. In addition, the context, namely Kelner's wearing of combat fatigues and holding of a gun, strongly suggests that he was making a threat. The fact that, unbeknownst to his listeners, Kelner never intended to carry through with the threat is irrelevant.
In addition, even if Arafat did not see the television broadcast, the telecast was reasonably calculated to reach him because of the large audience and public dissemination of the threat. At the very least, the threat would reach the United States government which would be forced to take precautions to protect the foreign dignitary while in New York. These sorts of precautionary expenses are just the sort of things that are sought to be prevented by punishing threats.
The defendant might argue that he did not intend to threaten, but rather was just using rhetorical hyperbole. This argument likely will fail. First, in contrast to Watts, Kelner intended his speech to reach his target, Arafat, rather than to energize his supporters with strong rhetoric. In addition, unlike Watts there was not a jovial atmosphere surrounding his comments. Furthermore, even if there was a degree of hyperbole in Kelner's remarks, he certainly knew there was a substantial risk that his statements would be viewed as intimidating or frightening by Arafat.
The actor prong is met since the action was threatened to be undertaken by the defendant's co-conspirators. The defendant stated that "we are planning to assassinate Mr. Arafat" and that his people were out "now" to carry out the deed. In addition, the fact that the defendant was holding a gun while speaking suggests his own willingness to act.
Finally, the reasonable listener prong is met because a reasonable person would likely interpret Kelner's statement as being a serious threat. A specific target was named as well as a specific time and place for the carrying out of the assassination. In context, this will likely be viewed as a serious threat, for the reasons discussed above. Thus, under the proposed test Kelner's speech would not be protected.
This outcome seems right because Kelner's speech will cause disruption, fear of an assassination attempt, and possibly coercion. In addition, Kelner had ample alternative ways to convey his message that the Jews would protect themselves against Arafat and the Palestinians without making a direct threat.
To test my theory further, consider the following variation on Kelner: In this version the defendant is speaking at a rally and says "supporters of Arafat oughta have their necks broken." This might be protected under the more protective Kelner standard which requires proof that the threat would be viewed as likely to be carried out imminently, but might be considered a threat under the reasonable listener standard of the other circuits. This hypothetical situation is similar to that of Claiborne Hardware where such a speech was found protected by the Supreme Court. It would also be protected under my proposed test.
The intent prong would probably be met since the speaker appears to intend to intimidate Arafat supporters. However, there are two possible reasons why it might not be met. First, depending on the locale and coverage of the rally it is possible that there was little or no likelihood that Arafat or his associates would receive the threat, which would cut against this prong. Second, it is possible that given the context of the speech, the words could be viewed as hyperbole used as a rhetorical device. In this hypothetical, the speaker is rallying support from the crowd much as the defendant in Watts did as well as Evers in Claiborne Hardware.
Whether or not the intent prong is met, the actor prong is not met since there is no suggestion that the speaker or his associates would act. If we were given more information, such as that the speaker and his associates were stockpiling weapons or organizing an army, or if there had been violence committed by the speaker or his associates, then the outcome would be different. As long as the speaker did not recklessly suggest that he or his associates would act, the speech will not meet the actor prong.
Under my modified reasonable listener prong, the reasonable listener test would not be met since the speech, given at a public rally, was not directed at a specific target, but rather at a group much as in Claiborne Hardware. Even though individual members of a group may feel threatened, the lack of specificity makes it less likely that the threat is serious.
Since this speech fails the actor prong and the reasonable listener prong it would not be considered a true threat. Thus, the only way that the speech could be found unprotected if my proposed test were in place is under the Brandenburg analysis for incitement of illegal activity. This outcomes fits with the holding of Claiborne Hardware and with the First Amendment's protection of political advocacy and allowance of strong rhetoric even where violence is advocated.
Conditional threats are a variation of the direct threat where the threat is conditional on some event or action. Conditional threats are often used to coerce the victim to do something against his will. Going back to the example of Kelly, the spurned girlfriend, imagine her calling up Randall and saying "If you don't get back together with me, I'll use you for target practice." All courts would find this threat unprotected unless, given the context, it was said clearly in jest. For all the reasons stated above under the direct threats analysis, this would be unprotected speech.
The intent prong is clearly met since there is no doubt that Kelly wished to intimidate and coerce Randall. Similarly, it is obvious that the actor prong would be met since Kelly said she would be the actor. The reasonable listener prong also would be met since a reasonable person would have no trouble identifying Kelly's words as a threat. The conditional nature of the statement would not make the statement protected because the statement is intended to directly coerce Randall to act differently than he otherwise would.
By contrast, the conditional nature of the speech in United States v. Watts was related to whether or not the defendant would serve if drafted -- it was not tied to the actions of another person whom he wished to coerce. Punishing Kelly's speech here makes sense because despite the conditional language, her statement will frighten Randall, possibly coerce him to act against his free will, and may force both Randall and the police to take disruptive precautionary measures.
Under my proposed test the conditional speech in Watts would still be protected. The intent prong would most likely not be met. The defendant's intention was not to intimidate or frighten the President but rather to make a political statement to his audience. This intention was fairly clear from the context. Watts made the alleged threat for purposes of rhetorical hyperbole rather than to intimidate, frighten, or coerce the President beyond what is allowable political advocacy.
Even though it is likely that the President or the Secret Service would be apprised of Watts' speech, since it was given in a public forum near the White House, this is not enough on its own to satisfy the intent prong. One could argue that the defendant was at least reckless that his speech would be viewed as a threat, but there is no evidence in the record that he suspected the President would be threatened. Given his goals, he seems at most negligent.
The actor prong is met because the defendant was going to be the one to act. It is possible that the reasonable listener prong could be met, but it is not likely since the statement was not taken seriously by the listeners, and was conditioned on an event the speaker vowed would never occur.
Because the defendant did not purposely, knowingly, or recklessly wish to intimidate or coerce the President, and a reasonable person would not view the statement as threatening, the speech in Watts would be protected. This is the right outcome not only because it is consistent with the holding in Watts, but also because it promotes open debate and vigorous advocacy.
In State v. Chung, the defendant told several of his fellow teachers that he planned to kill the school principal. The Hawaii Supreme Court held that his statements were punishable threats even though the defendant did not convey his threat directly to the principal. The outcome would be the same under the proposed test.
The intent prong is clearly met since the defendant explicitly told two other teachers that he wanted to and planned to kill the principal. In addition, the teachers were likely to warn the principal of this threat. Even though the speech was to a third party it was directed to a specific target and there was no suggestion that the speaker was using hyperbole to make a rhetorical point, such as drumming up support for a teachers' strike.
The actor prong is also met since the defendant said that he would be the one to do the killing. Finally, the reasonable listener prong is met since a reasonable person hearing the statement would have concluded that the defendant's statement was a serious threat. There was a specific target named and there was no suggestion that the speaker was joking. Holding the defendant here responsible for his threat even though it was told to a third party makes sense because his speech has little or no value and causes all of the harms sought to be prevented by the true threats exception to the First Amendment.
Warning threats are one crucial area where so-called threatening speech should be protected. Current tests, as discussed above, leave open the door to punish warning threats. At the same time, one does not want to overprotect this speech and allow truly threatening speakers to couch their threat as a warning and thereby escape punishment. The following scenarios will help illustrate the difference by clarifying how warning threats can be distinguished under the proposed test.
First, consider United States v. Dinwiddie. The defendant, Mrs. Dinwiddie, allegedly threatened an abortion provider, Dr. Robert Crist, and other staff at a Planned Parenthood clinic. One of the defendant's statements could be interpreted as a warning rather than as an actual threat. Specifically, Mrs. Dinwiddie yelled the following through a bullhorn: "Robert, remember Dr. Gunn [a physician killed a year earlier] . . . . This could happen to you . . . . He is not in the world anymore . . . . Whoever sheds man's blood, by man shall be shed." The defendant then went even further and made a clear threat by saying "You haven't seen violence yet until you see what we do to you." The District Court found that, given their context, both Mrs. Dinwiddie's statements constituted punishable true threats.
The outcome would be the same under the proposed test. The intent prong is met since the defendant purposely or at the very least knowingly wished to intimidate and coerce her victim to stop providing abortions. In addition, Dr. Crist unquestionably received the threats since she spoke to him directly. Mrs. Dinwiddie's history of advocating violence against abortion providers and personally attacking Planned Parenthood employees lends support for the proposition that she purposely intimidated and frightened Dr. Crist with threat of violence.
The one possible argument in the defendant's favor is that her statements were rhetorical hyperbole. However, unlike in Watts and Claiborne Hardware her speech was not given to a large audience of supporters, but rather was conveyed directly to the victim. There was no evidence that she was trying to whip up the support of her listeners. One could argue that her efforts to convince just one individual, Dr. Crist, constituted some degree of advocacy, and therefore she is entitled to use hyperbole even in this more private context. However, a more restricted application of the factor of rhetorical hyperbole should be taken in order to protect the rights of threatened individuals. Where there is not a large audience, it is harder to argue that the speech is harmless advocacy rather than a threat against the specific listener who is targeted. Furthermore, the speaker is free to address an individual for the purpose of advocacy so long as the speaker does not threaten the listener with violence or substantial property damage.
The actor prong is also met here since she mentioned that "we" were going to take action. Even if Mrs. Dinwiddie had only made her first statement regarding Dr. Gunn's death, it is likely that the actor prong still would be met because there are considerations outside the explicit words Dinwiddie used which suggest that she was, at the very least, reckless with regard to conveying the idea that she or her associates would act. After all, Dinwiddie was a known advocate of violence, had repeatedly threatened clinic workers, and had signed a petition supporting the murder of the very man she told Dr. Crist to remember.
In addition, Mrs. Dinwiddie directly targeted and threatened Dr. Crist. This is not the sort of general threat against a group that suggests that the speaker and her associates will not act. Any test for true threats cannot require a defendant to explicitly say that she is personally going to injure the victim because such a test would allow crafty speakers to escape punishment by carefully choosing their words. As the above analysis shows, the context of the speech and the background of the speaker will differentiate between thinly veiled actual threats and legitimate warning threats.
The reasonable listener prong would also be met by Dinwiddie's statements since a reasonable person would take the defendant's statements seriously. A specific target was identified and a reference was made to past violence. There is no evidence that the statement was made in jest and the defendant made repeated threats to her victim. Unlike in Watts, there was no unlikely condition required before the act took place; unlike in Claiborne Hardware, the message was conveyed directly to a specific target rather than to a group primarily composed of supporters.
It makes sense to punish Dinwiddie's speech because her statement causes all of the harms that threats can produce, and her right to advocate her political views does not outweigh the individual's right to be free from fear and intimidation. Furthermore, she still can present her views without directly threatening an individual doctor.
Now consider the following variation on Dinwiddie: a Catholic priest speaks at an anti-abortion rally outside a Planned Parenthood clinic. There are known supporters of violence in the audience, but the priest has always stated that he opposes violence. The priest warns the doctors that their lives are in danger from angry anti-abortionists if they continue to perform abortions.
The priest's speech fails the actor prong of the test. This sort of speech is a warning threat and should be allowed even if the priest hopes to discourage Dr. Crist and others from performing abortions. It is possible that the priest does intend to frighten Dr. Crist with the foreshadowing of violence, but given his nonviolent stand it is unlikely that he is suggesting that either he or his associates would carry out such a threat. Therefore, the priest cannot be punished under the proposed test.
The actor prong is the key difference between this hypothetical and Dinwiddie. In Dinwiddie, there was evidence that the defendant, a known advocate of violence, was threatening the victim she would act. In addition, the priest threatened abortion doctors in general, not a specific individual doctor. This more diffuse threat suggests that the alleged threat is not a true threat.
It is important that the priest's speech not be restricted in this instance, because this is the kind of strong advocacy that the First Amendment serves to protect. The small danger that the priest or his associates might act must be balanced with the values behind the First Amendment of allowing vigorous debate even on controversial issues. As both Claiborne Hardware and Brandenburg demonstrate, the Supreme Court has protected speech even where violence is advocated and even where there is a backdrop of violence. To the extent that the priest's words might incite others to act, his speech should be analyzed under the Brandenburg test for incitement.
As suggested earlier, there is a risk that would-be threateners will couch their threats in speech carefully constructed to avoid implicating that they will act in order to avoid punishment under the proposed test. For example, an individual might have every intention of threatening an abortion provider but say "I personally wouldn't do anything nor would my associates, but others might" to avoid punishment with a careful choice of words. This problem will be avoided by looking to the speaker's intent and the factors that determine intent. If the speaker purposely, knowingly, or recklessly gives the impression that the speaker or the speaker's associates will carry out the threat, then the actor prong can still be met.
The determination of whether or not the speaker is, at a minimum, reckless as to the actor prong is related to the factors considered to show the speaker's intent. For example, if the speaker had committed acts of violence in the past, then it would be less believable for her to claim that she would not be the actor of the threat. Similarly, if she had made prior threats or told others of her intentions to commit a violent act against the victim her intention could be proven.
Consider another variation on Dinwiddie: What if the priest in the prior example, rather than warning the doctors about out of control anti-abortion activists, threatens with the wrath of God? This in fact happened not only in Dinwiddie, but also in Simpson v. Burrows, in which the court in dicta suggested that threats of God's vengeance might be enough on their own to subject the speaker to punishment.
Even though a reasonable listener might be frightened, intimidated, and even coerced by such threats of God's vengeance, these are clearly threats or warnings which should be protected. For example, people who believe atheists, homosexuals, and adulterers are sinners who are going to be struck down by the hand of God should be able to voice their beliefs. The line is crossed, however, when the speaker suggests that he or his associates will help God by taking action down on Earth. But until that line is crossed the speech must be protected. This again shows why the actor prong is essential to any test for determining a true threat.
A similar analysis should be applied to protect threats of voodoo and witchcraft. Since voodoo is essentially considered a religion it should be viewed similarly to a threat of God's vengeance. In addition, the supernatural elements will most likely mean that the threats will fail the reasonable listener requirement. But even if the alleged threat meets the reasonable listener prong, it will fail the actor prong since God or another supernatural force is the implied actor rather than the speaker or his conspirators.
Some of the most difficult cases to analyze are those where the threat is not explicit. The recent controversy over the Nuremberg Files and the Jake Baker e-mails falls into this category of implied threats. The current tendency is to punish speech which I believe should be protected under the First Amendment. Because there are no clear and consistent guidelines for determining a true threat, courts have the flexibility to decide cases based on individual interpretations and instinct. The courts' reliance on subjective factors often results in decisions which are more restrictive of speech than is allowed by the First Amendment. A few illustrative examples demonstrate that the proposed test will more fairly protect and, in some instances, punish speech.
Recall that the Nuremberg Files were published on a website which listed names of abortion supporters and crossed out the names of abortion doctors after they were murdered. The fact that the jury in Oregon found this speech threatening and awarded a large judgment makes clear that that a reasonable person likely would view the website as constituting a serious threat. In fact, many scholars agree with this outcome and believe that the Nuremberg Files should be punishable threats. The proposed test would reach a different conclusion.
The intent prong would likely be met because the crossing out of the doctors' names suggests that the website creators likely knowingly, or at least with reckless disregard, intimidated or frightened the doctors listed on the site with a threat of violence. Since the names were listed on a public website it is reasonable to say that the targets were likely to receive the threat.
The actor prong, however, would not be met since there was no evidence that the website creators or their associates threatened that they would take any action beyond cataloging the names of abortion providers and identifying when they were injured or killed. Even though there was a backdrop of violence, there was no evidence that the website creators had acted violently in the past or had any plans to take action themselves against the doctors.
The posting does not meet the minimum recklessness standard of the actor prong because nothing beyond the broader backdrop of anti-abortion violence suggests that the website creators were aware of a substantial risk that the site implied that they would take action themselves. The website creators may well have been negligent as to whether others would believe they were threatening that they would act, but that is not a sufficient mental state for the actor prong. The fact that the website may have facilitated or incited violence against abortion doctors is another issue altogether.
The reasonable listener prong is met since a reasonable person reading the website, given the context of abortion clinic violence and the crossing out of the names, might well view it as a serious threat. This is especially true since nothing about the site suggested that it was a joke. Thus, the Nuremberg Files would be protected speech, under the proposed test, only because it fails the actor prong.
The validity of this outcome is shown, in part, when one considers the mirror websites which I discussed earlier. Several First Amendment advocates put up mirror sites which exactly replicated the Nuremberg Files. No court has evaluated whether these mirror sites constitute true threats, but the publishers and/or servers of these sites have all since dismantled them. Under the proposed test, the mirror sites would be protected.
First, the intent prong would not be met. There is no question that the intent of the mirror websites was to allow access to the files for scholars and the public to examine, as well as to make a point about free speech. The individuals who posted the mirror sites clearly had no intention of threatening the doctors, and many of them were in fact supportive of abortion rights. However, if nothing about the website suggested who the webmasters were or their motives, one might conclude that the mirror website creators were reckless as to the threatening impact of the site.
The actor prong also would not be met since the publishers of the mirror sites were not interested in taking any action and nothing in the sites suggested they or their associates would take any action. The reasonable listener prong would likely be met since a reasonable listener, absent a disclaimer in the website, would find the site to pose as equally serious a threat as the original site. Thus, because the mirror sites at the very least fail the actor prong, and probably the intent prong as well, such sites would be protected speech under the proposed test.
These mirror sites should be protected in the interest of scholarship and public access to information. Even though the District Court of Oregon did not protect the original Nuremberg Files, they should also be protected speech. One reason for this is that it is difficult to justify allowing mirror sites with identical postings while banning the original website, without further evidence that the original creators were going to act, simply because of the website publishers' differing positions on abortion. The proposed test would eliminate the potential for such value-based judgments. Instead, the crucial issue would be the actor prong and whether the website creators were at a minimum reckless as to whether they suggested that they or their associates would carry out the alleged threat.
As discussed in detail earlier, the actor prong safeguards speakers who are using strong rhetoric for the allowable intention of coercion and intimidation as long as they do not threaten that they or their associates will take action. Even though protecting the Nuremberg Files is a controversial outcome for many people, it is the right result because it protects heated political advocacy where there is not sufficient proof that the speakers threatened that they would take any action. The proposed test results in a fairer, more viewpoint neutral evaluation of speech and protects social commentary, opinion, and advocacy where there is no evidence that the speaker purposely, knowingly, or recklessly implied that he or his associates would carry out any threat. In addition, banning speech like the Nuremberg Files will severely limit anti-abortion speakers' ability to express harsh condemnation of abortion because such strong speech against the backdrop of violence will often seem threatening to the reasonable listener unless one considers an actor prong.
Another case which had a strong propensity to be incorrectly decided under the current test is United States v. Fulmer, which I described at some length above in Part IV. Briefly, the defendant found new evidence in a case he had asked the FBI to investigate. He called the FBI agent and told him "the silver bullets are coming." The FBI agent felt threatened and the defendant was convicted at the initial trial.
The jury decision in Fulmer shows that using only a reasonable speaker test, a reasonable person might find the defendant's words to be a threat even though there was clear evidence that the defendant did not intend to make a threat. This evidence was not information that the FBI agent would have had and thus it is very possible that a reasonable person in his shoes also would be frightened.
However, it is unjust to punish the defendant in this instance for poorly chosen words. He used the phrase the "silver bullets are coming" as a way of suggesting impending justice -- this was a phrase his acquaintances heard him use often and understood to be nonthreatening speech. The proposed test reaches the correct conclusion here where the current test fails.
The intent prong is not met since there is clear evidence that the defendant did not purposely, knowingly, or even recklessly frighten, intimidate, or coerce the FBI agent. At worst, he was negligent. The actor prong might not be met because it is not clear who would act given that the phrase "the silver bullets are coming" uses a passive construction and is vague at best. However, if the speech itself were considered threatening it likely would appear that the defendant was threatening that he would act because of his initiation of the phone call. As discussed above the reasonable listener prong is probably met.
Fulmer highlights the importance of adding an intent prong to the test for true threats because without such a requirement, inarticulate and merely negligent speakers would be unfairly punished.
The highly publicized Jake Baker case involved a series of e-mails sent between Jake Baker and an e-mail pen pal, Arthur Gonda, who was never located. There was also one story which was posted publicly on the web which described in explicit detail the rape, mutilation, and murder of a woman who had the same name as one of Baker's classmates at the University of Michigan. The e-mails, which were privately exchanged and not accessible via the web, contained explicit discussions of both correspondents' interests in harming and sexually abusing women. Even though the Court of Appeals and District Court both believed the e-mails were protected speech under the First Amendment, there was a vociferous dissent on the Court of Appeals for the Sixth Circuit and several requests for an en banc hearing.
First let us consider the e-mails: Baker and his pen pal discussed their interest in harming women, but they never conveyed such threats to their targets and there was no evidence that they took any act towards doing so. With no other protection than the reasonable listener test, the defendant in Baker could have been convicted since the e-mails might have been interpreted as serious threats - certainly they were not conditional nor mere jests.
Under my proposed test Baker's e-mails would be protected. The intent prong is not met since it is difficult to imagine how Baker purposely, knowingly, or recklessly intimidated, frightened or coerced anyone given the privacy of the communication. This emphasizes one of the key factors in addressing this prong, likelihood of receipt; even if the prosecutors could identify a target of Baker's threat, it is highly unlikely that the victim would be apprised of the threats.
The actor prong is met since in each e-mail Baker discusses action that he will take. The reasonable listener prong is not met since the lack of specificity of a target suggests that a reasonable recipient would not be threatened by the communication. Baker mentioned an interest in attacking girls in his dorm, and local thirteen and fourteen year old girls, but none by name. Baker could possibly be charged with conspiracy to commit sexual assault or other crimes but certainly not for making a true threat. This outcome is appropriate because the First Amendment protects private correspondence, thoughts, and fantasies even if the contents of such communication are disturbing or repulsive.
The posted story is another matter. If the government had pursued the conviction for the story, this speech might not have been so easily protected. The story as posted on the web was titled with his classmate's last name. The story repeatedly used her first and last names. The story identified her as a classmate of Baker's and described in painful detail, in the first person, the narrator and an accomplice sexually molesting her, raping her, burning her, and then finally setting her on fire.
Absent Baker's classmate's name this would be protected speech -- nothing more than a gruesome fantasy. Baker's use of his classmate's name moves the speech into the category of unprotected speech under the proposed test. First, the intent prong is met. Even though Baker may not have purposefully intended to intimidate his classmate, he would certainly have known that if she read the story she would be intimidated by it, given its gruesome and explicit nature. It is also likely, since Baker posted the story on a public website and used his classmate's name as the title, that the victim would receive the threat.
The actor prong is met as well since the story describes a first person account of the brutal torture and Baker is her classmate and knows her whereabouts. Because of the realistic tenor of the story, and the use of the first person, Baker was at least reckless as to the implication that he would be the one to carry out the threatened attack.
The reasonable listener prong is also met since a reader of the story would certainly be threatened by it if she were in the classmate's shoes. Nothing suggests that the story was a joke, and Baker's choice to use his classmate's name in a public forum takes the speech out of the world of fantasy and gives it a specific target who would be and was in fact deeply affected by the story. This result seems right because the speech being restricted is of limited value and the harms are very great for the victim. In addition, Baker is free to post such a story without using his classmate's name.
1. On Target's Property
When a cross is burned on a person's property, it is a strong signal that a direct threat has been made against that person. Courts in several circuits have found that the burning of crosses on a person's private property is an unprotected threat. Similarly, under the proposed test burning a cross on the private property of the intended victim would not be protected.
First, the intent prong is met -- by placing the cross and then setting fire to it, the defendant is making a clear statement that he purposely wishes to intimidate, frighten, or coerce the inhabitants of the property. The fact that the cross is burned on the victim's property makes it seem like the speaker promises to follow through with further violent action or destruction of property. Furthermore, it is highly likely that the inhabitants of the property would notice a burning cross on their front lawn.
The actor prong is met here since the simple fact that a person has no qualms about breaking the law by trespassing and burning a cross suggests that the cross-burner may engage in further violent action against the victim. This is different from the Nuremberg Files and Claiborne Hardware because the threat was personally conveyed by the speaker to the intended target and the speaker took physical action against the target by burning the cross which emphasizes the speakers willingness to take further action.
The reasonable listener prong is also satisfied. This is especially true given the historical meaning of cross burning and would be even clearer if the targets were African-American. By burning the cross on an individual's property, the speaker violates the victim's right to be free from threats. This countervailing right outweighs the speaker's right of expression. This outcome makes sense because there are alternatives ways the speaker could communicate his message of racial antipathy, including burning a cross in a different location.
2. On Public Property or Consenting Person's Private Property
Cross burning on public property is another story. This starts to move more into the category of public and political speech that the First Amendment generally protects. Racist individuals should be able to express their racist beliefs by burning crosses on their own property and even in some circumstances on public property. The proposed test would reach the same result as the Eighth Circuit in United States v. Lee, which held that when a cross is burned on property not belonging to the victim, further evidence is required before a defendant can be convicted of making a threat, including evidence that the defendant intended to threaten his victims. To the extent that cross burning incites violence against blacks or other groups, it should be analyzed under Brandenburg, not under a theory of threats.
The main issue here is what is the actual threat allegedly being made by the cross burning. The burning of the cross can imply a threat against blacks in general, but the implication here is not as clear as in the situation where a cross is burned on someone's property. By burning a cross on public property there is no suggestion that violence or substantial property damage will be committed against anyone in particular. Racists are allowed to burn crosses and protestors to burn flags without being punished for threatening an individual or group of individuals. In keeping with these principles, the proposed test would allow cross burning at rallies.
Specifically, the intent prong might not be met depending on the evidence of whether any particular effort was made to intimidate, frighten or coerce the victim(s). As discussed above, it is not as clear that violence or property damage is promised by the burning cross since the cross is not on a particular target's lawn. The main audience is likely in this instance not to be local blacks, but rather supporters of the cause. It is difficult to show intent where there is no clear victim.
In addition, it might be difficult to assess whether the targets of the alleged threat actually received the threat, depending on the location of the burning. If the cross burning was in the town square of a small town and anti-black rhetoric was involved, one might say the threat was directed towards all of the African Americans in town. In a small town, all of the blacks would likely hear about the cross burning. In a large city, it is possible that the African American community might not hear about the cross burning or it might be in such a remote place that no one besides members of the group burning the cross would find out about it. Of course, if the cross burning received news coverage that would be a different story.
The actor prong would probably be met since the speakers are actively burning a cross. This action suggests that the speaker and his associates will have no qualms about taking further action. The reasonable listener prong is probably not met because there was no specific target beyond perhaps blacks in the community, and the nature of the threat is unclear. Because the reasonable listener prong cuts against finding a threat, a jury may conclude that no threat was made.
It makes sense to allow cross burnings which are not targeted at specific victims because such speech allows individuals to express their opinions and release their frustrations without threatening anyone with violence. To the extent that such cross burnings may incite violence, they are more properly analyzed as incitement rather than as a threat to a specific group such as African Americans. It is important to be tolerant of such "group threats" in order to allow speakers to express their views.
To confirm that this is the right result, consider an event which occurred in 1988 at Gallaudet University, the nation's only liberal arts college for deaf and hearing impaired students. The students wanted a deaf president of the university to be hired. When the Board of Trustees chose yet another hearing president, the students protested and burned an effigy of the new president on the campus grounds. The intent of the burning effigy may well have been intimidation, but it was also clearly political advocacy since the whole protest was in the context of a civil rights movement for the deaf. The students' goal was not to take violent action against the president, but rather to make a political statement which enabled them to release their anger and frustration.
This kind of speech should be protected by the First Amendment. Allowing burnings of effigies and even crosses allows people to strongly and dramatically advocate their views. Where no specific individual or individuals are threatened such speech must be allowed.
A more difficult question arises if the cross is burned across from the NAACP headquarters. The intent prong is more likely to be met here since the speakers chose the location in an effort to intimidate NAACP members and visitors. However, the reasonable listener prong might not be met since there is no indication that the cross burning threatens any future acts of violence or destruction of property. Even though the target is more specific than in the burning of a cross on a generic patch of public property, there is no threat made to harm individual members or visitors to the NAACP. It seems that the burning of the cross across from the NAACP would need to be accompanied by some additional actions or statements before it was banned. Otherwise, it seems if anything less directed and specific a threat than the burned effigy of the specific individual at Gallaudet.
In United States v. Hart, the defendant parked two Ryder trucks in the driveways at an abortion clinic and was convicted for making a threat in violation of the Freedom of Access to Clinic Entrances Act (FACE). The defendant was a known anti-abortion activist. Even though the defendant's purpose could have been merely to interfere with the clinic's daily functioning, it was clear to the court that he had threatened the clinic.
Given the circumstances in the actual case, the speaker would also be convicted under the proposed test. The intent prong is met since the defendant admitted that he wanted to threaten the abortion clinic. It is clear that the speaker purposely or at least knowingly intimidated and frightened the clinic by choosing Ryder trucks, which he admittedly knew were the ones used in the Oklahoma City bombing. In addition, there was no doubt that the clinic staff was aware of the defendant's actions given the location of the trucks.
The actor prong is also met since the defendant parked the trucks himself. The implicit threat was that the trucks had bombs in them. Since the defendant parked them, he was at the very least reckless as to the fact that others would view him as the one to carry out the threatened action of detonating the explosives believed to be contained within the trucks.
The reasonable listener prong is certainly met since a reasonable person would conclude that a reasonable person in the clinic would be intimidated and fearful especially against the backdrop of the Oklahoma City bombing which the clinic workers remembered and associated with the trucks. This is the right result because the defendant's intentional action caused all of the harms associated with a true threat, and he had other effective ways of expressing his anti-abortion views.
In August of 2000, the "Late Late Show" with Craig Kilborn flashed the words "snipers wanted" under a picture of Republican presidential nominee George W. Bush. Given the comic nature of the show and the surrounding presidential election campaign, this speech is clearly protected by the First Amendment. However, under only a reasonable listener test, the sense of humor of the jury would determine whether or not Kilborn's speech is punished.
The proposed test would afford clearer protection for this speech. The intent prong would not be met. It seems likely that Kilborn could prove that his intent was merely to convey a joke. Bush might even have viewed the call for snipers as a joke himself. At worst, Kilborn was probably negligent as to the risk that his speech would be viewed as a threat.
The best protection for Kilborn, however, is the actor prong of the test. Kilborn and his associates had no intention of acting themselves since they were asking for snipers. Nor would anyone take seriously Kilborn's alleged effort to hire snipers to carry out the assassination. The reasonable listener prong would most likely not be met given the context of the show and Kilborn's history of telling jokes on the show. Given this background most, although certainly not all, listeners would have thought the segment was a joke. The only way to assure protection for this sort of parody is to have more than just a a reasonable listener test. Kilborn's call for snipers should be protected in order to allow public commentary and open channels for parody and humor.
Most of us can recall times when we have muttered, under our breath, things like "I'll get you" or "I could kill you" or "If only I had a gun" to people who are infuriating us. These off the cuff expressions are statements which clearly should not be punished as true threats. Not only are these natural outbursts, but they also provide a safety valve which allows people to express their frustration without resorting to violence.
One can imagine a variation on Lovell v. Poway, a case where a student sued after her school suspended her for allegedly threatening her guidance counselor. In our hypothetical, the student is criminally charged for her outburst. The evidence in the case showed that the student, Lovell, had been shuttled back and forth between administrative offices for hours while unsuccessfully trying to change her course schedule. Lovell was clearly at the end of her patience by the time she met with the guidance counselor. When the counselor told her that she might not be able to make the requested changes Lovell said "I'm so angry, I could just shoot someone." Lovell apparently immediately apologized for her outburst.
Under only a reasonable listener test, Lovell could have been found guilty of making a criminal threat. The court in the actual case was especially moved by the backdrop of violence in schools today: "In light of violence prevalent in schools today, school officials are justified in taking very seriously student threats against faculty or other students."
This seems like the wrong outcome because Lovell was expressing exasperation, not a violent threat. This is just the sort of letting off steam that is considered in the safety valve justification for allowing some forms of incendiary speech. By expressing anger, an individual's rage is defused before it can build over time and result in future violence. This is true even if the enraged individual has no present interest in committing a violent act. Under the proposed test, Lovell could not be seen as having made a true threat. She did not purposely, knowingly, or recklessly frighten, intimidate, or coerce the counselor. She was exasperated and had a verbal outburst which both parties agree she immediately apologized for. Again the importance of having an intent prong is evident.
Another example of a spontaneous outburst was in United States v. Crews, where a patient in a mental ward threatened to kill President Reagan after watching "The Day After" on TV. The court upheld a conviction for threatening the President. This outcome seems wrong given both the impetus for the speech and the context of the speech.
The defendant would not have been convicted under my proposed test. First of all, the statement appears to be an outburst much like Lovell's and therefore does not meet the intent prong of the test. There was little reason for the speaker to believe that his comment, made essentially to the TV screen, would be viewed as a threat. In addition, it was unlikely that anyone would report his statements to the authorities.
The actor prong and reasonable listener prong might be met since the speaker suggested he would act and a jury might find that given the mentally unstable character of the speaker and the content of the threat it should be taken seriously. However, since the intent prong is not met, the defendant's speech would be protected under the proposed test.
This is the right outcome because it allows for free expression where there are few harms. There was little doubt that the defendant was not about to carry out any action against the President. Even though the Secret Service was called in, only one officer investigated and he left after a short interview with Crews. In addition, the President was not likely to even be apprised of such a minimal risk and certainly he would not be disrupted or coerced in any way by this threat. Even though it is important to vigilantly protect the Chief Executive of our country, it is also important to allow for vocal criticism of the President as well as angry outbursts, as long as a serious threat has not been made.
With the exception of Crews above, cases in which the President's life was threatened would come out the same under the proposed test. For example, in United States v. Roy, a private in the Marines who was awaiting transport from Camp Pendleton to Vietnam heard that the President was visiting the Marine base the next day. He and his fellow officers joked about shooting the President. The defendant decided to call the operator from the pay phone in the barracks. He told the operator to tell the President not to come to the base because if he did he was "going to get him." At trial, despite the defendant's claim that he was joking, he was convicted of threatening the President. The Ninth Circuit affirmed his conviction.
The outcome would be the same under the current test. The intent prong is met. Even though the defendant might not have purposely threatened the President, he certainly did so knowingly. It was likely that the operator would convey his threat to the authorities, especially since the President was visiting the next day.
The actor prong is also met since the defendant stated that he would be the one to kill the President. The reasonable listener prong is also met given the circumstances of the speech: the President was to arrive the next day, the speaker would most likely have been seen as a disgruntled draftee being shipped off to Vietnam, and the operator took his threat seriously. This outcome makes sense because of the directness of the threat, the great jeopardy to our nation when the President's life is in danger, and the potential disruption the threat could cause during the President's visit.
It has long been the tradition in this country to allow dissent and protests. Boycotts, protests, and strikes are prime examples of situations where the First Amendment stands guard over the rights of speakers. The Supreme Court has repeatedly emphasized that protests and labor strikes are protected speech. To the extent that the definition of what constitutes punishable threats is vague there is a danger of suppressing and chilling those who wish to protest because they will censor themselves to avoid punishment. Limits on speech during boycotts and protests will hamper the ability of civil rights leaders, social leaders, and politicians to communicate and rally support for their causes. It is therefore of the utmost importance to apply the proposed test to boycotts, protests and strikes.
Perhaps abortion protests and labor strikes seem like two unlikely bedfellows, but in a discussion about threats they go hand in hand. In both protests and strikes there are excited participants who want to communicate their opinions and who strive to coerce either their employer or an abortion provider to change his behavior. In both cases there may be a backdrop or threat of violence.
In the context of abortion protests, rhetoric is heated and there is a history of violence from anti-abortion fanatics. However, abortion protesters have every right to speak out even in terms that seem incendiary. This was discussed earlier in some depth regarding the Nuremberg Files and United States v. Dinwiddie.
In order to assure the wide open debate on issues that the First Amendment and Brandenburg protect, some rhetoric advocating violence must be tolerated. However, speech crosses the line when a direct threat of violence is made against a specific target. This was the case in Dinwiddie and in United States v. McMillian where an anti-abortion activist threatened to shoot clinic staff and pantomimed shooting them by turning his hand into the shape of a pistol.
Similarly, workers have every right to threaten to strike, to strike, and even to advocate violence so long as it is not a true threat and is not intended to incite imminent violence and is not likely to cause imminent violence. This boundary was crossed in People v. Prisinzano during a labor strike at the Fulton Fish Market. The defendant approached some replacement workers and told them that once "the cops leave, the blood is going to run off of your bald fucking head," and "I'm going to get you." The court found that the defendant could be tried since his speech met the criteria for true threats.
The defendant's speech would also be punishable under the proposed test. The intent prong is met since the intent of the defendant was unquestionably to intimidate, frighten and coerce the replacement workers into leaving and to stop them from crossing the picket line. There was no question that the targets received the threat since they were spoken to directly. The actor prong is also met since the defendant threatened that he personally would act.
The reasonable listener prong is met because the threat was directed at a specific target. Even though the backdrop of the speech was a picket line, the defendant was not making a public speech and therefore there is no suggestion that he was exaggerating for rhetorical effect. Although one can certainly exaggerate in private speech as well, there is more need to protect hyperbolic public speeches as they contribute more substantially to the public debate on controversial issues. In contrast, individually conveyed rhetoric is more likely to be viewed as a threat than speech conveyed to a group.
For several reasons, Prisinzano comes out differently than NAACP v. Claiborne Hardware, where Charles Evers' speech during a boycott was found protected. First, the majority of Evers' speech was directly related to nonthreatening words about the importance of the boycott and nonviolence whereas in Prisinzano the defendant's only words to the alleged victims were direct threats of violence. In addition, the actor prong is met, unlike in Claiborne Hardware.
Furthermore, the speech in Prisinzano was directed to specific individuals, rather than to a group of boycott violators who may or may not have been present for Evers' speech. The less specific and immediate the target of the speech, the less it makes sense to consider it to be threatening violence. This difference is crucial because it shifts the test from my proposed test for true threats to the already established test for incitement under Brandenburg.
Even though protests and strikes are generally protected by the First Amendment, when protestors and strikers cross the line and make direct threats of violence, the mere political background of their speech does not make their threats protected. This outcome makes sense in Prisinzano because the defendant's speech caused all the harms that justify the true threats exception and he had ample room to speak without making threats. The defendant could even have advocated violence so long as he stayed within the requirements of Brandenburg, but he could not threaten specific individuals by suggesting that he would physically attack them.
Claiborne Hardware is one of the most difficult cases to analyze. The decision itself is fairly opaque about its basis for determining that Evers' speeches did not constitute true threats. Particularly troublesome is Evers' speech where he may have said "we're gonna break your damn neck." To the extent that Evers was this direct in his threat and a specific list was referred to with the names of boycott violators, the proposed test might punish Evers' speech even though the Supreme Court found it to be protected. However, if a less extreme interpretation of Evers' statement is applied in which there are no specific targets of the alleged threat, the proposed test would come out the same way as the Supreme Court which held that his speech did not constitute a true threat. The outcome under the proposed test would be in accordance primarily because of the actor prong.
Whether the actor prong is met depends very much on what Evers said. There is a dispute about the actual language he used and whether or not he was really conveying a true threat or merely voicing his disapproval of boycott violators. Assuming Evers never said "we" will break your necks, then the actor prong, which requires the threat to call for either the speaker or his associates to act, would not be met. In addition, even if he used stronger language, there is ample evidence that Evers was a strong believer in nonviolence and that the bulk of his speech emphasized nonviolence. Furthermore, there was no evidence linking Evers to any criminal activity. It was therefore unlikely that Evers suggested that he would himself act or that he would instruct others to commit violent acts. Thus, without more evidence the actor prong would not be met.
The intent prong might also not be met because Evers' statements seemed more directed at the audience of supporters and rallying the group than to threatening or intimidating anyone. Therefore, his speech was more rhetorical hyperbole than any sort of threat. However, Evers likely knew that if his speech was heard by potential boycott violators, they would be intimidated or coerced. Evers certainly wanted to coerce people to obey the boycott, but it is not clear that he threatened any acts of violence - without making such a threat, the intent prong cannot be met.
It is not clear whether the reasonable listener prong would be met. Several things suggest that it might not be. First, Evers spoke to a large crowd primarily composed of supporters who responded favorably to him. Much of the audience reacted with support rather than fear. Also, the threat was made to boycott violators in general, not to any specific people. However, because of the existence of the list of boycott violators, the reasonable listener prong could be satisfied.
Even if the intent prong and reasonable listener prong were met, Evers' speech would be protected because the speech fails the actor prong. The protection of Evers' speech, as interpreted here, makes sense because it preserves open and vehement debate and advocacy, which is crucial not only to the expression of ideas, but is a cornerstone of civil rights movements. However, if a similar situation arose in which an advocate of violence was known to have said "we're gonna break your neck" to a clearly identifiable group of individuals, the outcome should be different and such speech should be restricted. This type of speech would cross over from being a legitimate tool of advocacy to an unprotected true threat.
The absence of a clear and uniform test for determining when a true threat has been made presents a serious danger of punishing speech which should be protected under the First Amendment. The time for reform is now, not only because of the recent high profile Internet cases, but also because of new federal and state laws that are rapidly being developed that arise out of the law of threats. These include the increasing number of stalking statutes, anti-mask statutes, statutes prohibiting panhandling, and hate crimes laws. Each of these statutes has been developed as an extension of statutes on threats. Each statute focuses on specific behavior that individuals might find intimidating, coercive, or frightening.
In applying each of these statutes, the majority of the courts use the same sort of reasonable listener test to determine the speaker's First Amendment protections. More and more actions as well as speech will be restricted as threats without the protection of a clear and consistent test for determining the parameters of the First Amendment exception for true threats.
It is time for the Supreme Court to grant certiorari in a threats case to once and for all settle this murky area of law. Meanwhile, circuit courts should call for en banc hearings to settle inconsistent and unclear precedents within each circuit. State Supreme Courts also have an opportunity to step into the fray and clarify a more speech protective test under their individual state constitutions.
The courts should then, at a minimum, adopt Justice Marshall's recommendation in United States v. Rogers of including a subjective intent prong to the test for threats. Furthermore, as my proposed test suggests, the addition of an actor prong requiring the speaker or his associates to be the anticipated individuals to carry out the threat will protect warning threats and strongly worded political advocacy. Without such safeguards in place, speakers in both the political and private arena run the risk of being punished for speech that is misconstrued as a threat or which constitutes allowable advocacy. Until the test for true threats is reformed, I recommend that one watch one's words and one's websites carefully, or else one might find oneself on the wrong end of a threat -- not as a victim, but as an unexpected defendant.
 This hypothetical is based on Planned Parenthood of the Columbia/Willamette, Inc v. American Coalition of Life Activists, 41 F. Supp. 2d 1130 (D. Ore. 1999) [hereinafter Planned Parenthood III], Planned Parenthood of the Columbia/Willamette, Inc v. American Coalition of Life Activists, 23 F. Supp. 2d 1182 (D. Ore. 1998)[hereinafter Planned Parenthood II], and Planned Parenthood of the Columbia/Willamette, Inc v. American Coalition of Life Activists, 945 F. Supp. 1355 (D. Ore. 1996) [hereinafter Planned Parenthood I]. This case is more commonly referred to as the "Nuremberg Files" and is discussed at length in Parts IV and V.
 This hypothetical is based on United States v. Alkhabaz, 104 F.3d 1492 (6th Cir. 1997) and United States v. Baker, 890 F. Supp. 1375 (E.D. Mich. 1995) (note that the defendant changed his name while the case was on appeal). This case is commonly referred to as the "Jake Baker" case. It is discussed further in Parts III, IV, and V.
 This hypothetical is based on NAACP v. Claiborne Hardware, 458 U.S. 886 (1982), which I will discuss at greater length in Parts II, IV, and Part V. Note that Charles Evers, leader of the boycott in Claiborne County, was accused of saying "we're gonna break your damn neck" although the exact language used was disputed both at trial and on appeal. See discussion infra pp.12-13, 15-16, and note 55.
 This hypothetical is based on the United States v. Fulmer, 108 F.3d 1486 (1st Cir. 1997), which I will discuss in more detail in Parts IV and V.
 See John Rothchild, Menacing Speech and the First Amendment: A Functional Approach to Incitement That Threatens, 8 Tex. J. Women & L. 207, 240-41 (1999) (describing the greater impact of threats law on the Internet as opposed to on other mediums); Anna S. Andrews, When is a Threat "Truly" a Threat Lacking First Amendment Protection? A Proposed True Threats Test to Safeguard Free Speech Rights in the Age of the Internet, UCLA Online Institute for Cyberspace Law and Policy, at http://www.gseis.ucla.edu/iclp/aandrews2.htm (May 1999) (describing the proliferation of the Internet and the unique ease of conveying threats over the Internet); Melanie C. Hagan, Note, The Freedom of Access to Clinic Entrances Act and the Nuremberg Files Webs Site: Is the Site Properly Prohibited or Protected Speech?, 51 Hastings L.J. 411, 424-25 (2000) (describing the expansion of the Internet and its power to disseminate speech); Jeremy C. Martin, Note, Deconstructing "Constructive Threats": Classification and Analysis of Threatening Speech After Watts and Planned Parenthood, 31 St. Mary's L.J. 751, 779-80 (2000) (describing the unique potential for the Internet to be a vehicle to convey threats and harassment). See generally ACLU v. Reno, 929 F. Supp. 824, 831-44 (E.D. Pa. 1996) (describing the widespread use of the Internet and the ease of widely disseminating speech).
 The Nuremberg Files listed the names and addresses of abortion providers and then crossed off the names of doctors who were murdered. See Steven G. Gey, The Nuremberg Files and the First Amendment Value of Threats, 78 Tex. L. Rev. 541, 541-42 (2000). The District Court of Oregon found that the website could be considered an unprotected threat and a jury awarded a $107 million award. See id. at 541. This decision is currently on appeal in the Ninth Circuit.
 In the Jake Baker case, the Sixth Circuit found that private e-mails which described the sender's interest in committing gruesome and unlawful sexual acts were protected speech. See United States v. Alkhabaz, 104 F.3d 1492, 1496 (6th Cir. 1997). The court found that the e-mails did not constitute threats because the e-mails were private correspondence and therefore did not have a coercive goal. See id. The court did not address the legality of the defendant's posted story about the rape and murder of his classmate, because the government abandoned prosecution for the posted story prior to trial. See United States v. Baker, 890 F. Supp. 1375, 1380 n.6 (E.D. Mich. 1995).
 See John E. Nowak & Ronald D. Rotunda, Constitutional Law (5th Ed. 1995); Rodney A. Smolla and Melville B. Nimmer, Smolla and Nimmer on Freedom of Speech (3rd ed. 1996); Laurence Tribe, American Constitutional Law (2d ed. 1988).
 See, e.g., 18 U.S.C. § 115 (a)(1)(B); 18 U.S.C. § 248(a)(1); 18 U.S.C. § 844(e); 18 U.S.C. § 871; 18 U.S.C. § 875; 18 U.S.C. § 876. See also Franklyn S. Haiman, Speech and Law in a Free Society 217-18 (1970) [hereinafter Speech and Law] (describing various federal statutes related to threats).
 See, e.g., Cal. Pen. Code §§ 26, 76, 422; Fla. Statute § 836.05; Haw. Revised Statutes §§ 707-716; N.Y. Pen. Code § 240.25; see also Haiman, Speech and law, supra note 10, at 218.
 See, e.g., Tompkins v. Cyr, 202 F.3d 770 (5th Cir. 2000); Lane v. Cole, 88 F. Supp. 2d 402 (E.D. Pa. 2000); Simpson v. Burrows, 90 F. Supp. 2d 1108 (D. Or. 2000); Planned Parenthood II, 945 F. Supp. 1355 (D. Or. 1996); Wolfson v. Lewis, 924 F. Supp. 1413 (E.D. Pa. 1996); Cotton v. Duncan, 1993 WL 473622 (N.D. Ill. 1993); see also Kent Greenawalt, Speech, Crime, and the Uses of Language 291-292 (1989). See, generally, Note, First Amendment Limits on Tort Liability for Words Intended to Inflict Severe Emotional Distress, 85 Colum. L. Rev. 1749 (1985).
 I will not address in detail the justifications or values behind the First Amendment since many scholars have discussed this at great length. For those who wish to read some background information on the subject see Nowak & Rotunda, supra note 8; Smolla and Nimmer, supra note 8; Tribe, supra note 8. See also Greenawalt, supra note 11, at 16-34; Franklyn S. Haiman, Speech Acts and the First Amendment 7-9 (1993) [hereinafter Speech Acts]; Alexander Meiklejohn, Free Speech: And its Relation to Self-Government, 16-19 (1948); Wojciech Sadurski, Freedom of Speech and Its Limits 8-35 (1999); Gey, supra note 6, at 551.
 See R.A.V. v. St. Paul, 505 U.S. 377, 382-90 (1992).
 394 U.S. 705 (1969).
 See id. at 707.
 For a detailed discussion of why the two tests essentially are the same see discussion infra p. 19.
 458 U.S. 886 (1982).
 I borrow this term from Eugene Volokh who adopted the term from computer programmers as a way to evaluate any proposed test using tough scenarios to challenge the test's validity. See Eugene Volokh, Intermediate Questions of Religious Exemptions – A Research Agenda with Test Suites, 21 Cardozo L. Rev. 595, 599-600 and n. 8. Viewing a variety of factual situations allows the development of a test which works in a broad spectrum of situations regardless of the political or individual preferences of the scholar, jury, or jurist. See id. at 599-600.
 There is no need to punish threats of minor crimes since there is little harm from waiting to act until the illegal action is carried out. For a discussion of why only serious rather than minor crimes should be made criminally punishable as threats see Greenawalt, supra note 11, at 101. Greenawalt argues that one should not punish threats when the threat is not "a felony against a person or property, and the communication does not itself involve an abuse of authority." Id. at 103.
 The reader should note that I have limited my discussion primarily to threats of illegal action. Threats of legal action are also sometimes made illegal under blackmail and extortion laws where the speaker seeks compensation in exchange for keeping quiet about something or in exchange for the recipient taking certain action against his own will. Even though I use blackmail and extortion in my analysis of the policy beyond punishing threats, I do not include blackmail or extortion (except where the threat is of violence or destruction of property) in my proposed test or test suite because of the different statutes involved.
 The first three of these harms are listed in R.A.V. v. St. Paul, 505 U.S. 377, 388 (1992).
 See Arne Ohman, Fear and Anxiety as Emotional Phenomena: Clinical Phenomology, Evolutionary Perspectives, and Information – Processing Mechanisms, Handbook of Emotions 512-14 (eds. Michael Lewis & Jeanette M. Haviland 1993). See also Tompkins v. Cyr, 202 F.3d 770, 782 (5th Cir. 2000); United States v. Alkhabaz, 104 F.3d 1492, 1498 (6th Cir. 1997); Simpson v. Burrows, 90 F. Supp.2d 1108, 1121 (D. Or. 2000); Tompkins v. Cyr, 995 F. Supp. 664, 673-74 (N.D. Tex. 1998).
 See, e.g., Greenawalt, supra note 11, at 209 ("When direct threats to high government officials are made, extensive social resources are devoted to ascertaining whether a genuine danger exists and preventing actual attacks.").
 As the Third Circuit in United States v. Kosma, 951 F. 2d 549 (3d Cir. 1991), pointed out: "Not only is this statute meant to protect the President's life, but it is also meant to prevent the disruptions and inconveniences which result from the threat itself, regardless of whether there is any intention to execute the threat." Id. at 556.
 422 U.S. 35 (1975).
 Id. at 46-47 (J. Marshall, concurring).
 See e.g., Wolfson v. Lewis, 924 F. Supp. 1413, 1422 –23 (1996); see also infra note 29 regarding measures taken by abortion providers as a result of threats against their lives.
 See, e.g., Tompkins v. Cyr, 202 F.3d 770, 777 (5th Cir. 2000); United States v. Dinwiddie, 76 F.3d 913, 918 (8th Cir. 1996); Planned Parenthood II, 23 F. Supp. 2d 1182, 1186 (D. Or. 1998). In the Nuremberg Files case, the Federal Bureau of Investigations and United States Department of Justice offered twenty-four hour a day protection to the doctors on the Deadly Dozen List (a series of posters of twelve doctors offering a reward for information leading to arrest conviction and revocation of license to practice medicine). See id.
 See generally Steven J. Breckler, Emotion and Attitude Change, in Handbook of Emotions, supra note 22, at 464-5 (describing the coercive effect of threats). Kent Greenawalt calls such speech "situation altering" because it changes what the threatened party does from what they naturally would have done. See Greenawalt, supra note 11, at 68. For further discussion of coercive threats see id. at 92-94
 See Haiman, Speech and Law, supra note 9, at 210; see generally Jeffrey F. Webb, Political Boycotts and Union Speech: A Critical First Amendment Analysis, 4 J.L. & Pol. 579 (1988) (discussing the acceptance of pickets and strikes which have a coercive effect).
 See NAACP v. Claiborne Hardware, 458 U.S. 886, 910-11 (1982); accord Organization for a Better Austin v. Keefe, 402 US 415 (1971).
 See New York Times Co. v. Sullivan, 376 U.S. 254, 270 (1964) ("[D]ebate on public issues should be uninhibited, robust, and wide-open, and . . . it may well include vehement, caustic, and sometimes unpleasantly sharp attacks on government and public officials.").
 See Martin Luther King Jr., Letter From a Birmingham Jail, April 16, 1963 published on the web at www.mlkonline.com/jail.html. See also discussion infra at p.26-27.
 See United States v. Watts, 394 U.S. 705 (1969); Claiborne Hardware, 458 U.S. 886. The Court in Claiborne Hardware was explicit that speakers should not be required to turn "strong and effective extemporaneous rhetoric" into "nicely channeled" "dulcet phrases." Id. at 928.
 This is a brief though not complete summary of my test which will be presented in Part V.
 This list is primarily comprised of language from Watts which distinguishes a true threat from these other modes of speech. See Watts, 394 U.S. at 707-08.
 See Watts, 394 U.S. at 705.
 See id. at 706.
 See id. at 707.
 See id. 18 U.S.C. § 871(a) provides: "Whoever knowingly and willfully deposits for conveyance in the mail or for a delivery from any post office or by any . . . document containing any threat to take the life of or to inflict bodily harm upon the President . . ., the President-elect, the Vice President, or other officer next in the order of succession to the office of President . . . or knowingly and willfully otherwise makes any such threats . . . shall be fined not more than $1,000 or imprisoned not more than five years, or both." Id.
 See id. at 708.
 See id. at 708 (citing New York Times Co. v. Sullivan, 376 U.S. 254, 270 (1964)).
 Id. at 708.
 See id.
 The closest the Court came to clarifying what was meant was when it granted certiorari in the case of Rogers v. United States, 422 U.S. 35 (1975), "to resolve an apparent conflict among the Courts of Appeals" concerning the elements required for violating 18 U.S.C. § 871, the same statute the defendant in Watts was accused of violating by making a threat against the President. See id. at 36. The Court initially agreed to hear the dispute to determine whether or not a showing of the speaker's intent should be required to establish that a threat has been made. See id. However, after full briefing and argument, the Court found it unnecessary to answer that question since there were other grounds for reversing. See id. Justice Marshall in his concurring opinion, which Justice Douglas joined, criticized the Court for not clarifying the test for threats. See id. at 42-43 (J. Marshall, concurring). Marshall thought the test should require subjective intent on the part of the defendant. See id. at 44-48. I will discuss Marshall's concurring opinion in more detail in Part IV.
 458 US 886 (1982). I am not considering in my discussion threats of legal action which in certain circumstances have been found illegal by the courts, such as NLRB v. Gissel Co., 395 U.S. 575 (1969) (holding that employers could not threaten laborers that the factory would close if they striked unless this had basis in fact).
 See id. at 888. The boycott was called after the failure of white officials in Claiborne County to adequately address a list of demands from the black citizens of the county. The demands primarily focused on issues of racial equality and integration. The list of twenty-one needs included such requests as the desegregation of schools, employment of Negro policemen, equal treatment of customers, cessation of police abuse, desegregation of the courthouse, and the inclusion of Negroes on juries. See id.; NAACP v. Claiborne Hardware, 393 So. 2d 1290, 1295-97 (Miss. 1980).
 See Claiborne Hardware, 458 U.S. at 903-04; Claiborne Hardware, 393 So. 2d at 1297-98.
 See Claiborne Hardware, 458 U.S. at 903-04; Claiborne Hardware, 393 So. 2d at 1297.
 Of ten purportedly related incidents of violence the Supreme Court pointed to only four which demonstrated the "atmosphere of fear" and were "convincingly" determined to be retribution for breaking the boycott. These included: the firing of shots into the house of a boycott violator; the throwing of a brick through the windshield of a parked car owned by a boycott violator; the making of threatening phone calls to the home of another boycott violator; and finally the trampling of a flower garden owned by a woman who traded with boycotted merchants. Several other incidents of violence were reported but the Court was not convinced that they were related to the boycott. These other incidents included the stealing of a bottle of whiskey, a fist fight, and the pulling down of one man's overalls and spanking him for violating the boycott (this incident was only relayed via hearsay so the Court did not consider it as a related act of violence). In addition there was no evidence of any acts of violence after 1966, and therefore the violence had died down by the time Evers, in 1969, made the speech in which he allegedly said "we're gonna break your damn neck." See id. at 904-06.
 See id. at 890.
 Id. at 900.
 Id. at 938-39. (words in parentheses are audience responses). The exact words of this speech are known because it was recorded.
 Id. at 902. In contrast to the prior quoted speech, no one is sure exactly what was said because no audio recording was made of this speech.
 See Claiborne Hardware, 393 So. 2d at 1301. The Mississippi Supreme Court believed the entire boycott was illegal because of the backdrop of violence and threats, and therefore did not investigate in individual depth whether Evers' speech constituted true threats in and of themselves. See id. at 1299-1300.
 See Claiborne Hardware, 458 U.S. at 934.
 See id. at 929.
 Id. at 910.
 See id. 917-18, 925-26.
 See id. at 927.
 See id. at 928 n.71.
 Id. at 926-27.
 See id. at 928 (finding that a speaker is not required to turn "strong and effective extemporaneous rhetoric" into "nicely channeled . . . dulcet phrases").
 For a discussion of the history of the Supreme Court's and the United State's approval of boycotts and protests see ACLU Amicus in Claiborne Hardware at 7-8 and American Jewish Congress ["AJC"] Amicus at 6-7, 9-12.
 See Appendix to the Opinion, Claiborne Hardware, 458 U.S. at 934-40. The April 19th speech in which Evers made the comment that the sheriff could not sleep with boycott violators at night was recorded. Thus, we know both the content of the entire speech and the audience response. The comment was given in the context of a rally to energize the boycotters. In addition, this speech was made in the aftermath of the killing of a black youth by a white policeman which suggested an explanation for the heated rhetoric. See Cutler's Oral arguments in Claiborne Hardware at 16.
 See Petitioner's Brief in Claiborne Hardware at 8.
 The Supreme Court and other courts have recognized that how the audience of a speech reacts may be indicative of whether or not the speech was meant as a threat. The issue of whether the reaction of the recipient of threats should be admissible is discussed in more depth in Part IV.
 Recall that no incidents of violence related to the boycott were documented after 1966.
 Cutler, Oral Arguments in Claiborne Hardware at 17-18.
 See Memorandum of ACLU Foundation of Oregon Amicus Curiae, Planned Parenthood II, 23 F. Supp.2d 1182 (D. Ore. 1998), accessible at http://www.aclu-or.org/ppbrief.htm ("[A] 'true threat' must be directed against a specific and identifiable victim or victims. A threat directed generally against a group at large is not a 'true threat.' In fact, this distinction between focused threats that are directed against specific individuals and unfocused 'threats' that are directed against a large group generally, which in the heat of the moment can often be a part of the advocacy of ideas, may be helpful to understanding the Supreme Court's decision in NAACP v. Claiborne Hardware.").
 See Cutler, Oral Arguments in Claiborne Hardware at 14-15; see also Appendix to Petition for Certiorari in Claiborne Hardware, 66, 83, 111 , 184.
 Id. at 185.
 See AJC Amicus for Claiborne Hardware at 4 (discussing the obvious bias of the local Sheriff and the trial court).
 See Claiborne Hardware, 458 U.S. at 929.
 See Appendix to Petition for Certiorari at 194.
 See id. Evers was known for being a fervent follower of the teachings of Martin Luther King Jr. including King's strong belief in nonviolent means of producing change.
 In Brandenburg v. Ohio, 395 U.S. 444 (1969), a Klu Klux Klan ("KKK") leader was convicted under the Ohio Criminal Syndicalism statute for 'advocating . . . violence, or unlawful means of terrorism as a means of accomplishing industrial or political reform and for voluntarily assembling [to] advocate the doctrines of criminal syndicalism" Id. at 444-45. Specifically, Brandenburg led a KKK rally which was filmed by a member of the press. The film showed KKK members in robes and hoods, many of whom carried rifles. Brandenburg liberally peppered his speech with racial epithets against blacks and Jews. He also suggested that they should "Send the Jews back to Israel" and "Bury the niggers." Id. at 445-47. The Court held that the Ohio Syndicalism Act, under which Brandenburg was punished, was unconstitutional. See id. at 448-49. The Court articulated the Brandenburg test which still applies today. In order to punish someone for inciting unlawful activity, the test requires that the "advocacy is directed to inciting or producing imminent lawless action and is likely to incite or produce such action." Id. at 447.
 See Claiborne Hardware, 458 U.S. at 928. This seems strange since the Brandenburg test makes the fact that violence did or did not occur after a speech irrelevant. The test was about whether or not Evers' words were intended to incite imminent violence and whether or not they were likely to do so. The Court must have decided either that Evers was not likely to incite imminent violence or that he did not intend to do so, but the Court never makes clear on which basis Evers' speeches fail the Brandenburg test other than the lack of violence following his speech.
 Id. at 928.
 Id. at 929.
 For a discussion of circuit court confusion see ACLU Amicus in Planned Parenthood, supra note 71 ("[T]he Supreme Court . . . has not yet offered any extensive analysis of the 'true threats' doctrine."); Gey, supra note 7 (arguing that the Supreme Court has clearly articulated a test -- the Brandenburg test for threats -- but that the circuit courts have ignored it and have instead struggled to come up with their own, conflicting standards); Ashley Packard, Threats or Theater: Does Planned Parenthood v. American Coalition of Life Activists Signify that Tests for "True Threats" Need to Change?, 5 Comm. L. & Pol'y 235, 238, 248-49 (2000); Martin, supra note 5, at 756, 766; Leigh Noffsinger, Note: Wanted Posters, Bulletproof Vests, and the First Amendment: Distinguishing True Threats from Coercive Political Advocacy, 74 Wash. L. Rev. 1209, 1215-18 (1999); Robert Kurman Kelner, Note, United States v. Jake Baker: Revisiting Threats and the First Amendment, 84 Va. L. Rev. 287, 289 (1998).
 Even though it makes sense that people in as large and diverse a city as Los Angeles might view sexually explicit material differently than people in a more rural area, it makes no sense for people in two different cities or towns to view alleged threats differently -- unless a threat is somehow specifically contextualized to that place.
 108 F.3d 1486 (1st Cir. 1997).
 Id. at 1491-92.
 Id. at 1491-92.
 See id. at 1499-1500. See also United States v. Saunders, 166 F.3d 907, 913 (7th Cir. 1999); United States v. Dinwiddie, 76 F.3d 913, 925 (8th Cir. 1996); United States v. Orozco-Santillan, 903 F.2d 1262, 1265 (9th Cir. 1990).
 See United States v. Kosma, 951 F.2d 549, 557 (3rd Cir. 1991). In this case the Court stated that the reasonable speaker test "requires that the defendant intentionally make a statement, written or oral, in a context or under such circumstances wherein a reasonable person would foresee that the statement would be interpreted by those to whom the maker communicates the statement as a serious expression of an intention to inflict bodily harm upon or to take the life of the President, and that statement not be the result of mistake, duress, or coercion." Id.
 United States v. Roberts, 915 F.2d 889, 891 (4th Cir. 1990); accord United States v. Maisonet, 484 F.2d 1356, 1358 (4th Cir. 1973). However, the Fourth Circuit in United States v. Patillo suggests that where the speaker's threat is highly unlikely to reach the target, the courts should use a subjective standard in which the prosecution must show that the speaker had a present intention to injure, incite others to injure, or restrict the victim's movements. See United States v. Patillo, 438 F.2d 13 (4th Cir. 1971). See also discussion infra p. 24.
 See United States v. Daugenbaugh, 49 F. 3d 171, 173-174 (5th Cir. 1995); United States v. McMillan, 53 F. Supp. 2d 895, 904 (S.D. Miss. 1999).
 United States v. Miller, 115 F.3d 361, 363 (6th Cir. 1997) (holding that a true threat against the President can be made even when the speaker is incarcerated and unable to carry through the threat) (citing United States v. Smith, 928 F.2d 740, 741 (6th Cir. 1991)); United States v. Saunders, 166 F.3d 907, 912 (7th Cir. 1999) (quoting United States v. Khorrami, 895 F.2d 1186, 1192 (7th Cir. 1990)) ( " '[I]n a context or under such circumstances wherein a reasonable person would foresee that the statement would be interpreted by those to whom the maker communicates a statement as a serious expression of an intention to inflict bodily harm upon or to take the life of' another."); accord United States v. Glover, 846 F.2d 339, 344 (6th Cir. 1988); United States v. Cox, 957 F. 26 264, 266 (6th Cir. 1992) (holding that the reasonable person standard applies but making no distinction between the recipient versus speaker version of the test).
 United States v. J.H.H., 22 F.3d 821, 827-28 (8th Cir. 1994). The reasonable listener's evaluation is determined given the "entire factual context" and "whether the recipient of the alleged threat could reasonably conclude that it expresses 'a determination or intent to injure presently or in the future.'" United States v. Dinwiddie, 76 F.3d 913, 925 (8th Cir. 1996) (quoting Martin v. United States, 691 F.2d 1235, 1240 (8th Cir. 1982), cert denied, 459 U.S. 1211 (1983)). The Eighth Circuit also uses several factors to aid in the determination of whether a true threat was made. These include: 1) the reaction of the listeners, including the recipient of the threat; 2) whether the threat was communicated directly to its victim; 3) whether the threat was conditional; 4) whether the maker of the threat had made similar statements to the victim on other occasions; and 5) whether the victim had reason to believe that the maker of the threat had a propensity to engage in violence. See Dinwiddie, 76 F. 3d at 925; United States v. Hart, 212 F.3d 1067, 1071 (8th Cir. 2000). The court in Dinwiddie makes clear that the list is not exhaustive and that the factors do not constitute elements that must be met for the offense to have been made. See Dinwiddie, 76 F.3d at 925.
 See Roy v. United States, 416 F.2d 874,877-78 (9th Cir. 1969). The Ninth Circuit has held, as have most other circuits, that alleged threats must "be considered in light of their entire factual context, including the surrounding events and reaction of the listeners." United States v. Orozco-Santillan, 903 F.2d 1262, 1265 (9th Cir. 1990) (citing United States v. Mitchell, 812 F.2d 1250, 1255 (9th Cir. 1987)). There is some confusion in the Ninth Circuit about whether the subjective intent element has been added or not for several federal threats statutes. For a more in-depth discussion of the Ninth Circuit's treatment of these statutes see discussion infra pp. 24-25.
 United States v. Viefhaus, 168 F.3d 392, 396 (10th Cir. 1999).
 See United States v. Callahan, 702 F.2d 964, 965 (11th Cir. 1983); Lucero v. Trosch, 928 F. Supp. 1124, 1129 (S.D. Ala. 1996).
 See United States v. Adams, 73 F. Supp. 2d 2, 3 (D. DC 1999).
 534 F.2d 1020 (2d Cir. 1976).
 Id. at 1027. Although referring to 18 U.S.C. § 875(c), the court relied on the Supreme Court's holding in United States v. Watts, 394 U.S. 705 (1969), where the Court evaluated the parameters of the First Amendment exception for true threats under 18 U.S.C. § 871.
 See Kelner, 534 F.2d at 1027.
 See id.
 See id. at 1023.
 18 U.S.C. § 875 (c) provides that: "Whoever transmits in interstate commerce any communication containing any threat to kidnap any person or any threat to injure the person of another, shall be fined not more than $1000 or imprisoned not more than five years or both." There is no explicit mens rea requirement in this statute, however, all circuits have read in at least a general intent requirement. The general intent requirement merely requires that the accused know that he transmitted the statement in question and that he knew its contents. The defendant does not need to know or plan that the statement will be taken as a threat. See, e.g., United States v. Whiffen, 121 F.3d 18, 20 (1st Cir. 1997); United States v. Francis, 164 F.3d 120, 120-21 (2nd Cir. 1999); accord United States v. Davis, 876 F.2d 71, 73 (9th Cir. 1989).
 See Kelner, 534 F.2d at 1020-21.
 See id. at 1021.
 See id.
 See id.
 See id.
 See id. at 1022.
 See id. at 1021-22
 Id. at 1022
 See id. at 1028
 Id. at 1023.
 Id. at 1028 (emphasis added).
 See id. at 1024.
 The use of the Kelner test in United States v. Baker, 890 F. Supp 1375, 1381-82 (E.D. Mich. 1995) was not relied on by the Sixth Circuit in United States v. Alkhabaz which affirmed for different reasons the district court holding in Baker. See Alkhabaz, 104 F.3d 1492 (6th Cir. 1997)).
 See United States v. Francis, 164 F.3d 120, 122 (2nd Cir. 1999) (commenting that all circuits except for the Ninth Circuit have interpreted 18 USC §875 as being a general intent crime); see also United States v. Patillo, 438 F.2d 13, 15-16 (4th Cir. 1971) (holding that a showing of specific intent is required where there is little likelihood that a statement will be received by the threatened target).
 853 F.2d 676 (9th Cir. 1988)
 Id. at 680-81; see also United States v. King, 122 F.3d 808 (9th Cir. 1997) (holding that Twine is still good law and that the district court erred in not giving a specific intent instruction to the jury). The court in Twine highlighted the substantial harm present in a threat against the President that is "qualitatively different from a threat against a private citizen or other public official." The reason for this is that '[a] President's death in office has worldwide repercussions and affects the security and future of the entire nation." Id. at 681.
 See supra note 102.
 See King, 122 F.3d at 809.
 See, e.g., United States v. Davis, 876 F.2d 71, 73 (9th Cir. 1989).
 See United States v. Patillo, 438 F.2d 13,15-16 (4th Cir. 1971)
 See id.
 104 F.3d 1492 (6th Cir. 1997).
 See id. at 1495-96
 See id. at 1493.
 Id. at 1493. For the text of some of the e-mails see id. at 1499-1501 (Krupansky, J., dissenting) and United States v. Baker, 890 F. Supp. 1375, 1387-1390.
 See Alkhabaz, 104 F.3d. at 1493.
 See id. at 1493,1497 (J. Krupansky dissenting) (recounting text of story).
 See Baker, 890 F. Supp. 375 at 1380 n.6.
 See Alkhabaz, 104 F.3d at 1495. The District Court had relied on the Kelner test and decided that the threatened action was not imminent. See Baker, 890 F. Supp. at 1381-91. The Sixth Circuit did not rely on the Kelner imminence requirement in its holding in Alkhabaz. See Alkhabaz, 104 F.3d at 1496.
 See Alkhabaz, 104 F.3d. at 1495.
 See id. at 1496.
 108 F.3d 1486 (1st Cir. 1997).
 18 U.S.C. § 115(a)(1)(B) provides that whoever "threatens to assault, kidnap, or murder, a United States official . . . with intent to impede, intimidate, or interfere with such official . . . while engaged in the performance of official duties, or with intent to retaliate against such person on account of the performance of official duties . . . shall be punished" (emphasis added). Despite the use of the word intent in the statute, no court has interpreted this statute as requiring specific intent. The courts have decided that all that the statute requires is that person knowingly transmit a communication which is then viewed as a threat under the reasonable listener test. See, e.g., Fulmer, 108 F.3d at 1491; accord United States v. Davis, 876 F.2d 71, 73 (9th Cir. 1989).
 See Fulmer, 108 F.3d at 1489.
 See id. at 1489-90
 See id. at 1490
 Blacks Law Dictionary defines "misprision of felony" as "the offense of concealing a felony committed by another . . . ." Black's Law Dictionary 1000 (6th ed. 1990).
 See Fulmer, 108 F.3d at 1490
 See id. at 1490, 1499.
 Id. at 1490
 See id.
 See id.
 See id.
 See id. at 1503.
 The Court Clerk for the District Court of Massachusetts, Susan Tibo, confirmed that the defendant was found not guilty after his new trial.
 422 U.S. 35, 41 (1975).
 Id. at 44
 Id. at 47-48 (quoting New York Times Co. v. Sullivan, 376 U.S. 254, 270 (1964)).
 See id.
 See Noffsinger, supra note 82, at 1235. Also consider the story of one Walter Walker, told by Justice Douglas in his concurring opinion in Watts. Walter Walker was "a 15th century keeper of an inn known as the 'Crown,' [who] was convicted under the Statute of Treasons for telling his son: 'Tom, if thou behaves thyself well, I will make thee heir to the Crown.' He was found guilty of compassing and imagining the death of the King, hanged, drawn, and quartered." United States v. Watts, 394 US. 705, 709 (1969) (J. Douglas, concurring) (citing 1 J. Campbell, Lives of the Chief Justices of England 151 (1873)). Even under the reasonable speaker/listener test the speaker in this example might not be found guilty of threatening the King given the context, but the gruesome historical scenario is a clear reminder that words can be misconstrued. This highlights the need to protect speakers from the vagaries of the reasonable person test.
 See New York Times v. Sullivan, 376 U.S. 254, 279 (1964).
 See Brandenburg v. Ohio, 395 U.S. 444, 447 (1969). In Brandenburg, the Court emphasized the importance of distinguishing abstract teaching of the necessity for violence from the actual intent to prepare a group to commit violence. Punishment for the former violates the First Amendment. See id. at 448.
 See Sullivan, 376 U.S at 279-80.
 See Gertz v. Welch, Inc., 418 U.S. 323, 349-350 (1974). Even though it is true that compensatory damages can be received for negligent false statements of fact about private figures or even potentially any damages on a strict liability basis for false statements on a matter of private concern, the availability of criminal punishment for threats requires a higher threshold. See id.; Dun & Bradstreet v. Greenmoss Builders, 472 U.S. 749 (1985).
 See Gertz, 418 U.S. at 349-50.
 See discussion supra pp. 23-24.
 See United States v. Crews, 781 F.2d 826, 836 (10th Cir. 1986) (Logan, J., dissenting).
 The ACLU in an amicus brief filed in Planned Parenthood II argued that a subjective, speaker-based test be added to the threats analysis. Such a subjective test would "require evidence, albeit circumstantial or inferential, that the speaker actually intended to induce fear, intimidation, or terror, namely, that the speaker intended to threaten." See ACLU, Amicus Curiae, Planned Parenthood II, supra note 71, at *13. The ACLU points out that without consideration of the speaker's purpose "the risk is substantial that a speaker who did not intend to threaten, but merely intended to communicate an idea in exercise of First Amendment rights, could be criminally prosecuted or held liable for damages, including punitive damages, in a civil action. Because of that risk and the danger to the First Amendment values that it poses, a subjective, speaker-based test should also be required." Id.
The drafters of the Model Penal Code also include a requirement of specific intent in their definition of a terroristic threat which they define as threatening "to commit any crime of violence with the purpose to terrorize another or to cause evacuation or a building…or otherwise to cause serious public inconvenience, or in reckless disregard of the risk of causing such a terror or inconvenience." See Model Penal Code § 211.3 (emphasis added).
Steven G. Gey in his article "The Nuremberg Files and the First Amendment" argues that not only should a showing of subjective intent be made but that a complete Brandenburg analysis should be required. See Gey, supra note 6, at 595. As I discuss shortly, although I do not agree with the further addition of the Brandenburg test because it is overly protective of speech, Gey's addition of an intent requirement is a step in the right direction. See id. See also Andrews, supra note 5; Noffinsger supra note 82, at 1234.
 See e.g., United States v. Whiffen, 121 F. 3d 18, 21 (1st Cir. 1997) (making clear that it does not think a subjective standard should come into play in threats because it is not as protective of listeners).
 See Donald Alexander Downs, Nazis in Skokie: Freedom, Community, and the First Amendment 166 (1985) ("[T]he courts explicitly consider intent…in many vital areas of law, including constitutional law and First Amendment cases. The prosecution must demonstrate a defendant's specific intent to break a law in most serious criminal cases (acts mala in se)."); see also Noffsinger, supra note 82, at 1235 ("While impossible to know a speaker's thoughts and true intent, jurors should consider evidence of intent just as they would assess mens rea for any crime or cause of action requiring specific intent.").
 See e.g., United States v. Fulmer, 108 F.3d at 1499-1500 (describing the acceptance by other circuits of the admission of recipient's reaction testimony); United States v. Dinwiddie, 76 F.3d 913, 925 (8th Cir. 1996); see also supra footnote 87; Note, First Circuit Defines Threat in the Context of Federal Threat Statutes – United States v. Fulmer, 111 Harv. L. Rev. 1110, 1114 (1998).
 Fulmer, 108 F.3d at 1500.
 See First Circuit Defines Threat, supra note 171 (describing how despite the objective test adopted in Fulmer, the test really becomes a subjective one with the admission of reaction testimony).
 The eggshell skull doctrine allows plaintiffs to recover even though their injuries would not have occurred if not for their unique sensitivity.
 See discussion supra pp. 21-23.
 In Hess v. Indiana, 414 U.S. 105 (1973), the Court suggested that imminence requires that the action be likely to place within hours or at most in a day or two. For a criticism of the application of an imminence requirement to threats see Kelner, supra note 82, at 296.
 See discussion supra pp. 6-10.
 See United States v. Kelner, 534 F.2d 1020, 1029 (2nd Cir. 1976) (Mulligan, J., concurring).
 This recalls Charles Evers language used during the Claiborne County boycotts which the Mississippi Supreme Court held to be a true threat. See NAACP v. Claiborne Hardware, 458 US 886, 900-902 (1982); see also discussion supra pp. 12-13, 15-16.
 See New York Times v. Sullivan, 376 U.S. 254, 270 (1964); Brandenburg v. Ohio, 395 U.S. 444, 447-48 (1969).
 NAACP v. Claiborne Hardware, 458 U.S. 886, 900 (1982).
 See Gey, supra note 6, at 541-42. The original website was taken down as was a mirror site so there is currently no way to access the Nuremberg Files. The Wanted Ads had two forms. One was a poster called the Deadly Dozen which listed the names, addresses, and phone numbers of abortion providers and labeled them Guilty of Crimes Against Humanity and offered a $5000 reward for information leading to arrest, conviction, and revocation of license to practice medicine. See Planned Parenthood II, 23 F. Supp. 2d 1182,1186 (1998). The other was a poster of Dr. Crist which offered a reward to any ACLA organization that successfully persuades him to turn from his "child killing" ways. Id. at 1186-87.
 In total more than 200 names were listed. See id. at 1188.
 The Supreme Court in other areas has not allowed the imposition of liability for the publication of public information. See Gey, supra note 6, at 563-40; see, e.g., Florida Star v. B.J.F. ,491 U.S. 524 (1989). However, publication of unlisted phone numbers and unpublished home addresses garnered by following people home are not considered public information and have been routinely considered private matters by the courts. See United States Dep't of Defense v. Federal Labor Relations Auth., 510 U.S. 487 (1994); Whalen v. Roe, 429 U.S. 598, 599-600 (1977); Heights Community Congress v. Veterans Administration, 732 F.2d 526 (6th Cir. 1984); Painting Indus. Of Haw Market Recovery Fund v. USAF, 26 F. 3d 1479, 1486 (9th Cir. 1994). Of these cases only Whalen addresses First Amendment issues, but the cases do stand for the basic presumption that unlisted phone numbers and residential addresses are considered private.
 The rhetoric included such language as "'baby butchers' who deliver 'Satan' his daily diet of slaughtered babies'" and the site had lines of dripping blood. See Gey, supra note 6, at 555-56. However, as Gey points out graphics of blood dripping and aborted fetuses do not make the expression any more of a threat. See id.
 It is worth noting that mirror sites were put up by abortion rights advocates who were also strong free speech proponents after the Nuremberg Files were taken off the Internet. See Gey, supra note 6, at n.1. Could these mirror sites be considered threats since they contained the exact same information as the other sites? See discussion infra pp. 57-59.
 The District Court decision and the jury judgment are currently on appeal before the Ninth Circuit.
 This might be different if the murders occurred in the order in which the names were listed on the website. But even then, the speech might be considered protected.
 One of the defendants had been convicted of conspiring to bomb several abortion clinics, although it is not clear that he was involved in the website. One of the defendants stated after the release of the Deadly Dozen List that "It wouldn't bother me if every abortionist in the country today fell dead from a bullet." Planned Parenthood I, 945 F. Supp. 1355, 1362 (1996). The District Court also pointed to the background of violence in abortion protests. The court lists seven incidents in which abortion providers were either murdered or shot and injured. In none of these instances were any of the defendants involved. The court tried to connect the defendants to these violent acts by referring to the release of wanted posters, and speeches given by some of the defendants prior to the violence.
Consideration of a backdrop of violence in determining whether or not speech constitutes a threat is dangerous. Just because there has been violence over a number of years in a certain area this should not be sufficient to hold individuals liable. The Supreme Court made clear in NAACP v. Claiborne Hardware, 458 U.S. 886 (1982), that an individual who is a member of a group cannot be held accountable for violence committed by others. See id. at 920. Where the defendants are not connected directly to any violence, the analysis should fall under the Brandenburg test for inciting violence, not under a threat analysis. To this extent, Steven Gey is correct that the Nuremberg Files should have been analyzed using the Brandenburg test to determine whether the defendant's speech was intended to incite imminent violence and was likely to do so. However, as discussed at length in Part V, this is because the defendants were at most inciting violence, rather than threatening that they or their associates would act.
 Rankin v. McPherson, 483 U.S. 378, 387 (1987).
 See, e.g., Claiborne Hardware, 458 U.S. at 909-10.
 781 F.2d 826 (10th Cir. 1986).
 See id. at 829. The nurse alleged that the defendant stated, after taking a large dose of anti-depressants, that "If Reagan came to Sheridan, I would shoot him." Id. When later asked by a Secret Service agent what he had said, he denied saying that he would shoot the president and instead said that he had stated that it "would be in the best interest of this nation if that red necked, bigoted, war-mongering mother fucker were shot." Id. The patient's version of his statement sounds similar to the protected speech in Rankin v. McPherson, 438 U.S. 378, 380 (1987).
 See Crews, 781 F.2d at 826. Courts seem universally to accept greater restriction of threats against the president. See also Greenawalt, supra note 11, at 290-91 (describing the justification for punishing Crews even though if he threatened his sister rather than the President his speech should be protected).
 Crews, 781 F.2d. at 837.
 For the moment, I am only addressing the e-mails and not the posted story about Baker's classmate.
 See United States v. Patillo, 438 F.2d 13, 15 (4th Cir. 1971) (en banc); discussion supra p. 24.
 See Kelner, supra note 82, at 287, 289 ("[S]ince the line between 'threat' cases under Watts and 'incitement' cases under Brandenburg is ill-defined, courts occasionally mix and match cases from both lines of precedent as if they were interchangeable.").
 See Gey, supra note 6; see also Noffsinger, supra note 82, at 1209 (suggesting the application of a requirement that the statement pose a likelihood of imminent violence).
 See NAACP v. Claiborne Hardware, 458 U.S. 886, 927-29 (1982).
 Gey, supra note 6, at 558. Gey believed that the Nuremberg Files should be protected speech, but not because no threat was made, but rather because there was no proof that the defendants intended to commit acts of violence against the named individuals. Id. at 568-69.
 See discussion supra pp. 6-10.
 For example, Gey interprets the speech in Watts as protected because of the political nature of the speech, and Brandenburg as solely a test about political advocacy. Gey believes that the Nuremberg Files should have been protected speech because there was no imminent danger from the website. As my proposed test demonstrates, one can find the Nuremberg Files deserving of First Amendment protection without needing to add an imminence requirement.
Leigh Noffsinger argues that political advocacy should not be punished under a threats analysis, but rather under Brandenburg. See Noffsinger, supra note 82, at 1209. She argues that the first step in analyzing any statement as a threat is to separate out political speech from other speech. See id.
 See Haiman, Speech and Law, supra note 9, at 18 ("[T]he Supreme Court [has] declined to recognize a categorical distinction between political and private speech . . . .").
 See Paris Adult Theatre I v. Slaton, 413 U.S. 49, 84 (1973) (Brennan, J., dissenting) (citations omitted).
 Id. at 86-93
 See Miller v. California, 413, US 15, 24 (1973).
 See e.g., United States v. Baker, 890 F. Supp. 1375, 1390 (E.D. Mich 1995); see also Hagan, supra note 6, at 424-28 (describing ways of regulating speech unique to the Internet); Andrews, supra note 6 (suggesting that because of the unique nature of the Internet a new test for threats is warranted).
 See Baker, 890 F. Supp. at 1390.
 For example, Ashley Packard, supra note 82, suggests the application of eight factors, developed by Professor David Crump in an incitement context, to any statement to determine whether or not a threat has been made. These factors are: "(1) the express words or symbols uttered; (2) the pattern of the utterance, including any parts of it that the speaker and the audience could be expected to understand in a sense different from the ordinary; (3) the context, including the medium, the audience, and the seriousness of unlawful results, and whether they actually occurred; (5) the extent of the speaker's knowledge or reckless disregard of the likelihood of violent results; (6) the availability of alternative means of expressing a similar message, without encouragement of violence; (7) the inclusion of disclaimers; and (8) the existence or nonexistence of serious literary, artistic, political or scientific value." See id. at 261. It is difficult to imagine a jury being able to apply all of these factors in any coherent manner. Furthermore, factors are best used in the framework of an actual and clearly delineated test.
 Downs, supra note 170, at 156.
 See Rothchild, supra note 5, at 224, 231, 238.
 See Martin, supra note 5, at 776-77.
 See id. at 794.
 See discussion supra p. 19.
 See Martin, supra note 5, at 791-92.
 I use the Model Penal Codes formulations for these terms. "A person acts purposely with respect to a material element of an offense when …it is his conscious object to engage in conduct of that nature or to cause such a result." "A person acts knowingly with respect to a material element of an offense when . . . he is aware that his conduct is of that nature …. and if the element involves a result of his conduct, he is aware that it is practically certain that his conduct will cause such a result." "A person acts recklessly with respect to a material element of an offense when he consciously disregards a substantial and unjustifiable risk that the material element exists or will result from his conduct. The risk must be of such a nature and degree that, considering the nature and purpose of the actor's conduct and the circumstances known to him, it involves a gross deviation from the standard of conduct that a law-abiding person would observe in the actor's situation." Model Penal Code § 2.02.
 I call this the reasonable listener prong because as discussed earlier the reasonable speaker and listener tests boil down to the same thing – both focus on how a reasonable listener would react.
 If the target is not specific, the court should apply the Brandenburg test rather than a threat analysis to determine if the speaker's words were unprotected incitement of violence. This is the approach the Court took in NAACP v. Claiborne Hardware, 458 U.S. 886, 927-29 (1982); see also discussion supra p. 17.
 For a discussion of the value and origins of using a test suite, see supra note 18.
 See United States v. Fulmer, 108 F.3d 1486 (1st Cir. 1997); discussion supra pp. 21-25.
 See discussion supra pp. 32-38.
 See Bose Corp. v. Consumers Union of United States, Inc., 466 U.S. 485, 505 (1984); see also Eugene Volokh and Brett McDonnell, Freedom of Speech and Independent Judgment Review in Copyright Cases, 107 Yale L. J. 2431, 2431-32 (1998).
 466 U.S. 485 (1984).
 Independent appellate review is required when analyzing fighting words and incitement, which are analytically similar to threats as exceptions to the First Amendment. The Court in Bose describes the fundamental policies behind having independent appellate review when limiting speech:
"[T]he Court has regularly conducted an independent review of the record both to be sure that the speech in question actually falls within the unprotected category and to confine the perimeters of any unprotected category within acceptably narrow limits in an effort to ensure that protected expression will not be inhibited. Providing triers of fact with a general description of the type of communication whose content is unworthy of protection has not, in and of itself, served sufficiently to narrow the category, nor served to eliminate the danger that decisions by triers of fact may inhibit the expression of protected ideas. The principle of viewpoint neutrality that underlies the First Amendment itself, also imposes a special responsibility on judges whenever it is claimed that a particular communication is unprotected."
Bose Corp., 466 U.S. at 505
Independent appellate review has been applied to fighting words, incitement, obscenity, negligent publication of criminal solicitation, government employee speech, commercial speech, content-neutral speech restrictions, and other reviews of cases arising under the First Amendment. See Hurley v. Irish-Am. Gay, Lesbian & Bisexual Group, 515 U.S. 557, 567 (1995); Luke Records Inv. v. Navarro, 960 F.2d 134, 138 (11th Cir. 1992); Herceg v. Hustler Magazine, Inc. 814 F.2d 1017, 1021 (5th Cir. 1987); Yakubowicz v. Paramount Pictures Corp., 536 N.E.2d 1067, 1071 (Mass. 1989); Braun v. Soldier of Fortune Magazine, Inc., 968 F.2d 1110, 1120-21 (11th Cir. 1992); Swineford v. Synder County, 15 F.3d 1258, 1265 (3rd Cir. 1994); Peel v. Attorney Registration & Disciplinary Comm'n, 496 US 91, 108 (1990); Revo v. Disciplinary Bd, 106 F.3d 929, 932 (10th Cir. 1997); Turner Broad. Sys., Inc. v. FCC (Turner I), 512 U.S. 622, 666-67 (1994); Association of Community Orgs. for Reform Now v. St. Louis County, 930 F.2d 591, 595-96 (8th Cir. 1991).
 Volokh and McDonnell, supra note 224, at 2432.
 See Watts v. United States, 394 U.S. 705, 707 (1969).
 United States v. Kelner, 534 F.2d 1020, 1020-21 (2nd Cir. 1976); see also discussion supra pp. 21-23, 31-32.
 See Kelner, 534 F.2d. at 1028.
 See id. at 1021.
 I specifically chose to tone down the rhetoric by using the phrase "oughta have their neck broken" rather than "we're gonna break your damn neck" which Evers in Claiborne Hardware was accused of saying. I think this is a fairer comparison because of the dispute over what was said and the larger context of Evers' speech. See discussion supra pp. 13-16.
 Compare with Lucero v. Trosch where the defendant Father Trosch, a Roman Catholic Priest, threatened to kill a fellow guest on the Geraldo show who was an abortion provider. However, when questioned by Geraldo he admitted that he would not kill him, he merely thought he should be dead. Thus, the court found his words protected even if Dr. Lucero felt threatened by him because he clearly stated that he had no intention of acting and his threats were simply hyperbolic advocacy. See Lucero v. Trosch, 928 F. Supp. 1124 (S.D. Ala. 1996) Again a Brandenburg analysis might apply here, although such an argument is unlikely to succeed since there was no suggestion of imminence or likelihood of action given the context.
 This is a different question than whether or not the defendant would actually be drafted and then forced to fight. The fact that the speaker believed that the event on which he conditioned his threat would never occur suggests that he was not serious about the threat.
 75 Haw. 398 (1993).
 See id. at 403-405.
 See id. at 403
 See discussion supra pp. 32-38.
 76 F.3d 913 (8th Cir. 1996).
 See id. at 917-18.
 Id. at 925.
 See id. at 926.
 See id. at 925-26. The defendant also had signed a petition in support of Michael Griffin, an anti-abortion activist who killed Dr. David Gunn. See id. at 918.
 See factors listed under intent prong, supra pp. 44-45.
 90 F. Supp. 2d 1108 (D. Or. 2000).
 See id. at 1130.
 The topic of hate speech is hotly debated in the legal community. Without complicating the discussion of threats, it is worth noting that there is a certain degree of overlap. Hateful speech damning homosexuals or Jews to perdition is not likeable nor even something one would seek to encourage, but a society that allows free expression of ideas and the separation of church and state must tolerate even repulsive speech. Furthermore, the suppression of hateful speech about groups that are favored may prevent critical speech at a later date about groups who we think should be criticized. For example, telling a member of the KKK that God will damn him for his bigotry and anti-Christian values should be allowed. See generally, J.S. Mill, On Liberty, Clear and Present Danger: The Free Speech Controversy 24-26 (Ed. Nicholas Capaldi 1969) (discussing the importance of allowing all opinions no matter how controversial or untrue to freely flow). For a discussion of whether hate speech is protected by the First Amendment see Haiman, Speech and Law, supra note 9 at 27-34.
 This is consistent with the case law which has not allowed threats of voodoo curses to be a basis for asylum in this country. See, e.g., Adebis v. INS, 952 F.2d 910 (5th Cir. 1992). But see ACLU Says School Barred Girl for Magic, L.A. Times, October 28, 2000, at A11 (describing a lawsuit in the US District Court in Tulsa, Oklahoma on behalf of a 15 year-old high school student who was suspended for casting a spell on her teacher which allegedly caused the teacher to fall ill).
 See discussion supra at pp. 34-37.
 See, e.g., Martin, supra note 5, at 781-83; Rothchild, supra note 5, at 223, 231 (although treating them as menacing speech); Andrews, supra note 5; Hagan, supra note 5, at 444. But see Gey, supra note 6, at 543-44.
 As discussed earlier, the web creators could be happy that abortion doctors were killed. In fact one of the web creators stated that when he crossed out a name on the website, it was a way of saying "I told you so." David Rovella, Site Author Notes when Pro-Choice Doctors Become Fatalities. Are the Judges Next? Nat'l L.J., Nov. 23. 1998, at A6. See also discussion supra pp. 28-31.
 It is unlikely that the imminence requirement of Brandenburg would be met by the level of advocacy in the Nuremberg Files, however it is possible given the finding in Rice v. Paladin Enterprises, 128 F.3d 233 (4th Cir. 1997), cert. denied, 523 US 1074 (1998), that a charge of criminal facilitation might allow prosecutors to penalize the website creator. The doctors might also be able to file civil charges for invasion of privacy for the printing of private addresses and unlisted phone numbers. In other words, there may be other ways to prosecute or hold liable individuals who convey this type of speech, without forcing it to fit the requirements of true threats.
 See supra note 187.
 Hagan, supra note 5, at n.17. The site Gey refers to is no longer in operation. See Gey, supra note 6, at n.1. It is not entirely clear why the mirror sites were taken down, but it is certainly possible, even likely, that fear of large monetary judgments against those who posted mirror sites or servers who allowed them to be posted was a main incentive in taking down the websites.
 The mirror site creators' intent would be even clearer if they had put disclaimers at the beginning saying: "I am posting this duplicate of the Nuremberg Files solely in the interest of Free Speech."
 Consider that Adolph Hitler's Mein Kampf is banned in several countries, but is a protected form of speech in this country under the First Amendment. Hitler's book may generally pose a danger to Jews based on its content; however, there is little dispute that the availability of this book is useful for scholars and the public alike.
 See discussion supra pp. 32-38, 52-56.
 108 F.3d 1486 (1st Cir. 1997).
 See discussion supra pp. 26-27.
 See Fulmer, 108 F.3d at 1490.
 See id. at 1490-91.
 See United States v. Alkhabaz, 104 F.3d 1492, 1493 (6th 1997).
 See id.
 See id.
 See id. at 1496 (Krupansky, J., dissenting).
 See Alkhabaz, 104 F.3d at 1504 n.10 (Krupansky, J., dissenting); United States v. Baker, 890 F. Supp. 1375, 1388 (E.D. Mich. 1995).
 Even a charge of conspiracy would be difficult to make stick since there was no overt act, and it's not so clear from the e-mails that any agreement, even an implied one, was made.
 See Alkhabaz, 104 F.3d. at 1497.
 See id.
 See id.
 See id. at 1498.
 Of course if Baker took off his classmate's name, but then suggested to that classmate that the story was about her, the story might be even more clearly a threat than it is in its original form.
 See United States v. Hayward, 6 F.3d 1241 (7th Cir. 1993); United States v. Juveniles J.H.H., 22 F. 3d 821 (8th Cir. 1994).
 The burning of a cross in the United States historically symbolizes the Ku Klux Klan's systematic intimidation and lynching of blacks in the name of white supremacy. See generally John Hope Franklin and Alfred A. Moss Jr., From Slavery to Freedom: A History of African Americans 385 (8th ed. 2000).
 See Haiman, Speech Acts, supra note 12, at 6 (suggesting that one could also distinguish burning of a cross on private and public property by pointing to the fact that rather than being speech, the burning of a cross on the victim's lawn is an act because it involves physical trespass and perhaps destruction of property which is legally punishable in and of itself).
 See, e.g., Texas v. Johnson, 491 U.S. 397 (1989) (holding that flag burning is protected by the First Amendment). But see Robin D. Barnes, The Reality of First Amendment Jurisprudence: Giving Aid and Comfort to Racial Terrorists, in Freeing the First Amendment: Critical Perspectives on Freedom of Expression 266-67 (David S. Allen and Robert Jensen eds., 1995) (treating public cross burning as racial terrorism).
 6 F.3d 1297 (8th Cir. 1993).
 See id. at 1304.
 See Texas v. Johnson, 491 U.S. 397 (1989).
 See generally Jack R. Gannon, The Week the World Heard Gallaudet, (1989).
 See id. at 79.
 See generally id.
 212 F. 3d 1067 (8th Cir. 2000).
 See id. at 1069; 18 U.S.C. § 248(a)(1).
 See Hart, 212 F.3d at 1069.
 See id. at 1072. If the defendant had obstructed the clinic's functioning, this would not be considered a threat. However, it would be punishable for trespass and interference with business.
 See id. at 1070.
 See id. at 1072.
 See L.A. Times, August 25, 2000, F22.
 This case was never prosecuted, but it is worth imagining a situation where it is.
 This outcome seems reasonable even now that George W. Bush is President. Kilborn's speech should still be protected under the proposed test. However, courts and juries are more likely to err in favor of punishing threats against the President if only the reasonable speaker/listener test is in effect.
 See Greenawalt, supra note 11, at 91 ("Since many threats are made in a flush of emotion that will dissipate and since sometime threats operate as a psychological alternative to immediate physical assault, the failure to make criminal all threats of serious harm is understandable . . .").
 See Smolla and Nimmer, supra note 8, at §1.04. ("It is thought that men will be less inclined to resort to violence to achieve given ends if they are free to express themselves through speech advocating such ends.") See also Martin Luther King Jr.'s Letter from the Birmingham Jail, supra note 180.
 90 F.3d 367 (9th Cir. 1996).
 See id. at 368.
 It is important to change the situation to one of a criminal indictment since there is a significant difference between speech that schools can discipline and speech that can be criminally punished.
 See id. at 369.
 See id.
 Id. In the actual case, the guidance counselor thought Lovell said: "If you don't give me the schedule change I'm going to shoot you." The district court judge believed Lovell's version of the events and so that is the version I use above. See id. at n.1.
 Id. at 369.
 Id. at 372. Even though it is true that this statement by the court was given in a case related to the disciplining of a student rather than in a criminal indictment, the strong language suggests a potential danger that under only a reasonable listener test the student might well be found guilty of making a criminal threat.
 781 F.2d 826 (10th Cir. 1986).
 See id. at 829-30; see also discussion supra p. 38-39.
 See Crews, 781 F.2d at 829-30.
 See discussion supra p. 38-39.
 416 F.2d 874 (9th Cir. 1969).
 See id. at 875.
 See id.
 See id.
 See id. at 876.
 See id. at 878-79.
 See, e.g., NAACP v. Claiborne Hardware, 458 U.S. 886, 909-10 (1982); see also McCalden v. California Library Association, 955 F.2d 1214 (1992) (holding that the making of a threat to protest the appearance of a Holocaust revisionist even if violence might break out was protected speech under the First Amendment); Haiman, Speech and Law, supra note 9, at 210. See generally, Webb, supra note 30. It is worth noting that a threat to protest would not even reach my proposed test because it is a nonviolent threat and does not threaten substantial property damage. As such the alleged threat would immediately be outside the scope of a true threat and could only be restricted under such other First Amendment doctrines as incitement analysis or a content neutral restriction based on time, place, and manner.
 For a discussion of the chilling effect of RICO on protestors see Brian J. Murray, Protesters, Extortion, and Coercion: Preventing RICO from Chilling First Amendment Freedoms, 75 Notre Dame L. Rev. 691, 694(1999). Murray warns that "the only people who can afford fully to exercise their right to protest are the wealthy, who can afford litigation, and the poor, who are judgment-proof anyway. Others, fearful of the costs of this sort of litigation – even if they ultimately "win" – will be intimidated into limiting the exercise of their rights to some level far below that which they are entitled under the First Amendment for fear of crossing the murky line from protest to trespass." Id. at 743-44.
 See discussion supra pp. 52-59.
 See discussion supra pp. 52-56.
 53 F. Supp. 2d 895 (S.D. Miss. 1999).
 See id. at 897.
 648 N.Y. S. 2d 267 (Criminal Court, NY, 1996).
 Id. at 527-28
 Id. at 537.
 See discussion supra and infra pp. 12-17, 32-38, 72-73.
 See discussion supra pp. 12-17, 32-38.
 See, e.g., 18 U.S.C. § 2261(a); Cal. Pen. Code §§ 422.6, 422.7, 646.9; Colo. Rev. Stat. § 18-9-111(4)(a)(II); Conn. Gen. Stat. § 53a-181; Ill. Crim. Code § 12-17.3(a); Mich. Compiled Laws § 750.411i (1994); N.J. NJSA § 2c:12-10; Or. Rev. Stat. §§ 163.732, 163.738, 163.730; City of Erie § 733.05 (struck down by the courts in KKK v. Erie, 99 F. Supp. 2d 583 (W.D. Pa. 2000); City Ord. No. 78 (1999) in Rev. Code of Indianapolis and Marion County §407-102. See also Haiman, Speech Acts, supra note 12, at 18-20 (commenting on the Second Circuit's upholding of a ban on panhandling in New York subway stations emphasizes that panhandling is a form of speech deserving of scrutiny under the First Amendment).