Harvard Law Review Harvard Law Review Harvard Law Review

First Amendment

Commonwealth v. Knox

True Threat Doctrine — Pennsylvania Supreme Court Finds Rap Song a True Threat.

While the Supreme Court in Watts v. United States1×1. 394 U.S. 705 (1969) (per curiam). placed “true threats” outside of the First Amendment’s protection,2×2. Id. at 707–08. the Court has never clearly defined what constitutes a true threat.3×3. See, e.g., Paul T. Crane, Note, “True Threats” and the Issue of Intent, 92 Va. L. Rev. 1225, 1226 (2006). The closest the Court has come to defining “true threats” was in Virginia v. Black, 538 U.S. 343 (2003), in which the majority characterized true threats as “statements where the speaker means to communicate a serious expression of an intent to commit an act of unlawful violence to a particular individual or group of individuals,” id. at 359 (citing Watts, 394 U.S. at 708). In seeking to explain the term, commentators have focused on the level of intent the First Amendment requires to establish a “true threat.”4×4. See Crane, supra note 3, at 1226. The main approaches are subjective tests (which require both general intent, meaning that the defendant knowingly made the statement, and specific intent, meaning that the defendant actually intended to make a threatening communication or to carry out the threat) and objective tests (which require only general intent). See id. at 1235, 1236, 1241. Recently, in Commonwealth v. Knox,5×5. 190 A.3d 1146 (Pa. 2018). the Pennsylvania Supreme Court sustained Jamal Knox’s conviction for terroristic threats and witness intimidation by concluding that he had a specific intent to threaten two police officers with his rap lyrics.6×6. Id. at 1158, 1161. While searching for specific intent is a step in the right direction, Knox demonstrates that a heightened mens rea requirement is insufficient to protect innocent speech if courts do not contextualize the allegedly threatening speech. Knox joins a growing series of cases in which rap, stripped of its artistic and political meaning, is put on trial. If the true threat exeception fails to treat rap music as art or to reconsider the emphasis on objective factors, the exception risks misapplication and disparate impact in First Amendment protection.

In April 2012, Jamal Knox was charged with multiple offenses after a routine traffic stop.7×7. Id. at 1148–49. While charges against Knox were pending, he “wrote and recorded a rap song entitled[] ‘F[uc]k the Police.’”8×8. Id. at 1149. To view the video, which features graphic pictures of victims of police brutality and lynching, see GhettoSuperStarTV, Ghetto SuperStar Committee ft Mayhem Mal & Soulja Beaz — Fuck the Police, YouTube (Dec. 2, 2012), https://www.youtube.com/watch?reload=9&v=3ScQEiBgzBo [https://perma.cc/72MM-ELUA]. The song was uploaded to YouTube, and the link was posted on a public Facebook page entitled “Beaz Mooga.”9×9. Knox, 190 A.3d at 1149. Trial evidence suggested the Facebook page belonged to Knox’s co-defendant and co-lyricist, Rashee Beasley. Id. at 1169. The lyrics described killing Pittsburgh police officers and informants and specifically referred to “Officer Zeltner” and “Mr. Kosko,” two members of the Pittsburgh police force who had participated in Knox’s arrest and were scheduled to testify against him.10×10. Id. at 1149. Some of the lyrics, as quoted by the court, included:
This first verse is for Officer Zeltner and all you fed force bitches/And Mr. Kosko, you can suck my dick you keep on knocking my riches/You want beef, well cracker I’m wit it, that whole department can get it/All these soldiers in my committee gonna fuck over you bitches/Fuck the, fuck the police, bitch, I said it loud. 
. . . .
So now they gonna chase me through these streets/And I’ma jam this rusty knife all in his guts and chop his feet/You taking money away from Beaz and all my shit away from me/Well your shift over at three and I’m gonna fuck up where you sleep. 
. . . .
They tunin’ in, well Mr. Fed, if you can hear me bitch/Go tell your daddy that we’re boomin’ bricks/And them informants that you got, gonna be layin’ in the box/And I know exactly who workin’, and I’m gonna kill him wit a Glock/Quote that.
Cause when you find that pussy layin’ in the street/Look at the shells and put my shit on repeat, and that’s on Jesus’ blood/Let’s kill these cops cuz they don’t do us no good/Pullin’ your Glock out cause I live in the hood/You dirty bitches, bitch!
at 1149–50.
An officer monitoring the Facebook page found the post and alerted Officer Kosko and Detective Zeltner.11×11. Id. at 1150. Knox was arrested and charged with two counts of terroristic threats and witness intimidation pursuant to sections 2706(a)(1) and 4952(a) of the Pennsylvania Crimes Code.12×12. Id. (citing 18 Pa. Cons. Stat. §§ 2706(a)(1), 4952(a) (2018)).

Following a bench trial, the trial court found that the song constituted a true threat, rejecting Knox’s argument that the song was protected under the First Amendment.13×13. Id. at 1151 (citing Trial Transcript at 462–64). The trial court also rejected Knox’s argument that he lacked the requisite mens rea because he was unaware the video would be posted online.14×14. Commonwealth v. Knox, Nos. 201206621, 201303870, 201304264, slip op. at 19–20 (Pa. C.P. Allegheny Aug. 11, 2015). The Superior Court affirmed in a memorandum opinion, but did not evaluate whether the song was protected speech.15×15. Commonwealth v. Knox, No. 1136, 2016 WL 5379299 (Pa. Super. Ct. Aug. 2, 2016) (mem.). Knox appealed.16×16. Knox, 190 A.3d at 1152.

The Pennsylvania Supreme Court granted review to address whether the rap video “constitute[d] protected free speech or a true threat.”17×17. Commonwealth v. Knox, 165 A.3d 887 tbl. (Pa. 2017) (per curiam). Writing for the majority, Chief Justice Saylor18×18. Chief Justice Saylor was joined by Justices Baer, Todd, Dougherty, and Mundy. held that the song was a true threat,19×19. Knox, 190 A.3d at 1161. rejecting Knox’s contention that the song was “merely artistic in nature” and “never meant to be interpreted literally.”20×20. Id. at 1153.

First, the court reviewed the doctrinal status of true threats, which the court noted had “its genesis” in Watts v. United States.21×21. Id. at 1155. After Watts, most courts used a variation of an objective test to determine whether a statement constituted a true threat (that is, would a reasonable listener or speaker interpret the statement as a true threat?), rather than examining the speaker’s subjective intent.22×22. Id.; see also Crane, supra note 3, at 1243. The Supreme Court, addressing true threats again in Virginia v. Black,23×23. 538 U.S. 343 (2003). rejected Virginia’s statutory presumption that cross burning implied an intent to intimidate, but failed to clarify whether subjective intent was required.24×24. Id. at 365–66; see also Crane, supra note 3, at 1259–60. Even so, the Pennsylvania Supreme Court read Black to mean that “an objective, reasonable-listener standard . . . is no longer viable.”25×25. Knox, 190 A.3d at 1156. The court drew two general principles from its review of true threat doctrine: first, that “the Constitution allows states to criminalize threatening speech which is specifically intended to terrorize or intimidate,” and second, that “whether the speaker acted with an intent to terrorize or intimidate” requires courts to consider “contextual circumstances such as those referenced in Watts.”26×26. Id. at 1158 (emphasis added). The Watts Court considered the content and nature of the statement, the setting, and listeners’ reactions, among other factors. Watts v. United States, 394 U.S. 705, 707–08 (1969) (per curiam).

Having found that true threat doctrine requires the speaker to have “acted with an intent to terrorize or intimidate,” the court then examined the “contextual circumstances” of Knox’s song.27×27. Knox, 190 A.3d at 1158. The court reviewed the content of the lyrics, finding that the song “d[id] not merely address grievances about police-community relations” or “include political, social, or academic commentary.”28×28. Id. Rather, the court held that the song “portray[ed] violence toward the police” and included personalized, “unambiguous threats.”29×29. Id.; see also id. at 1159 (describing the personalized nature of the threat). The court highlighted that Knox not only mentioned Detective Zeltner and Officer Kosko by name, but also referenced when the officers’ shifts ended.30×30. Id. at 1159. Additionally, the court determined that Knox was motivated by his recent arrest and Officer Kosko’s confiscation of Knox’s cash.31×31. Id. Finally, the court noted that the soundtrack included “bull horns, police sirens, and machine-gun fire ringing out over the words, ‘bustin’ heavy metal,’” elements designed to make the song more threatening.32×32. Id.

The court next addressed other contextual factors used in Watts and in Pennsylvania’s precedent, J.S. ex rel. H.S. v. Bethlehem Area School District,33×33. 807 A.2d 847 (Pa. 2002), abrogated by Knox, 190 A.3d 1146. to determine whether “the speech convey[ed] a serious expression of an intent to inflict harm.”34×34. Knox, 190 A.3d at 1159. The court found that the threats were mostly unconditional and that the officers had developed “substantial concern for their safety” as a result of the song.35×35. Id. The court also held that the officers had reason to believe that Knox might engage in violence because a loaded firearm had been found in Knox’s car.36×36. Id. at 1160. But, as the court noted, Knox was eventually acquitted of the firearm charges. Id. Although Knox did not send the video to the officers, the court concluded that there was no intention that the song “should not be conveyed to the police” once it had been uploaded.37×37. Id.

Finally, the court addressed Knox’s contention that rap songs frequently “include violent references” and “fictitious or fanciful descriptions of criminal conduct.”38×38. Id. Accepting that as true, the court found Knox’s song to be “of a different nature and quality,”39×39. Id. as Knox specifically called out two of the officers involved in his pending criminal case.40×40. Id. at 1160–61. Further, the court refused to insulate “an entire genre of communication” from legislative regulation under the First Amendment.41×41. Id. at 1161.

Justice Wecht, concurring in part and dissenting in part,42×42. Justice Wecht was joined by Justice Donohue. wrote to criticize the majority for failing to address “whether the First Amendment requires proof of specific intent.”43×43. Knox, 190 A.3d at 1162 (Wecht, J., concurring in part and dissenting in part). He saw the majority as having crafted only a “partial [test] that leaves uncertainty that will serve only to complicate and protract litigation in future cases.”44×44. Id. Justice Wecht, therefore, proposed a two-pronged approach that would require courts to examine first whether “reasonable recipients would consider the statement to be ‘a serious expression of intent to inflict harm’”45×45. Id. at 1165 (quoting J.S. ex rel. H.S. v. Bethlehem Area Sch. Dist., 807 A.2d 847, 858 (Pa. 2002), abrogated by Knox, 190 A.3d 1146). and second whether the speaker “specifically intended to intimidate” or threaten the recipients.46×46. Id. Justice Wecht maintained that only specific intent would suffice, finding that “[p]unishing statements that can be construed only as knowingly or recklessly uttered casts a net too wide.”47×47. Id. at 1164. Applying his own test, Justice Wecht nevertheless found that “Knox’s lyrics constitute true threats,” largely due to the personalization of the lyrics.48×48. Id. at 1171.

Much of the academic and judicial debate regarding the scope of true threats has turned on the requisite mens rea to separate protected speech from true threats. The Knox court, therefore, took a significant defendant-friendly step in claiming to require a finding of specific intent. However, the Knox court’s contribution in this regard was offset by its participation in a rising trend: Knox joins a series of cases in which rap — typically gangsta rap — is put on trial, stripped of its artistic and political meaning. Although it claimed to examine Knox’s subjective intent, the court ignored the artistic and political dimensions of his song and effectively applied an objective standard by giving undue weight to the listeners’ reactions. The court also relied on the personalization of the lyrics, allowing that factor to play an inappropriately determinative role. Knox demonstrates the need for courts to reconsider the factors weighed in determining subjective intent in order for “true threats” to proscribe only speech that conveys a serious intent to threaten.49×49. See Virginia v. Black, 538 U.S. 343, 359–60 (2003).

Knox joins a growing number of cases in which rap music is assessed by judges and juries who fail to treat it as art and instead regard the lyrics as literal and autobiographical. In their article Rap on Trial, Professors Charis Kubrin and Erik Nielson argue that the phenomenon is on the rise. They point to several cases in which the defendant’s rap music was used as evidence of violent intent.50×50. Charis E. Kubrin & Erik Nielson, Rap on Trial, 4 Race & Justice 185, 191–96 (2014). Kubrin and Nielson discuss the cases of Ronell Wilson, Clyde Smith, Anthony Johnson, Olutosin Oduwole, and Cameron D’Ambrosio in depth. Id. Most significantly, they note that prosecution training manuals recommend use of rap lyrics as potentially inculpatory evidence.51×51. Id. at 196; see also Donald Lyddane, Understanding Gangs and Gang Mentality: Acquiring Evidence of the Gang Conspiracy, U.S. Att’ys’ Bull., May 2006, at 1 (“In today’s society, many gang members compose and put their true-life experiences into lyrical form. . . . Law enforcement officials must remain mindful . . . to obtain inculpatory evidence in gang-related investigations and cases.”). Kubrin and Nielson suggest that the trend of using music lyrics at trial appears limited to rap alone.52×52. See Kubrin & Nielson, supra note 50, at 190. Separately, Professor Andrea Dennis has observed that she has identified only one instance in which defendant-authored lyrics at issue in a case were not rap lyrics.53×53. Andrea Dennis, Poetic (In)Justice? Rap Music Lyrics as Art, Life, and Criminal Evidence, 31 Colum. J.L. & Arts 1, 2 n.6 (2007) (citing State v. Koskovich, 776 A.2d 144 (N.J. 2001)). An independent search on Westlaw for lyrics specifically in true threat cases yields only rap lyrics. Although rap alone is prosecuted, other genres contain violent and graphic lyrics. See, e.g., Chelsea Crowell, Killer Songs: The 10 Creepiest Country Murder Ballads, Rolling Stone (Oct. 28, 2014, 1:38 PM), https://www.rollingstone.com/music/music-country-lists/killer-songs-the-10-creepiest-country-murder-ballads-151986/ [https://perma.cc/W9W3-HKV5].

Exacerbating the concern that prosecutors are singling out rap music is demonstrated bias in the perception of rap music — bias that may infect the judges and juries listening. In a recent study evaluating the effect of stereotypes of rap music on the perception of lyrics as “violent,” participants were asked to read violent lyrics from real country and folk songs.54×54. Adam Dunbar et al., The Threatening Nature of “Rap” Music, 22 Psychol. Pub. Pol’y & L. 280, 281 (2016). Some participants were told the lyrics were part of a rap song; others were told the lyrics were from a country song.55×55. Id. Participants who were told the lyrics were rap rated the lyrics as more offensive, literal, and in need of regulation compared with participants who were given identical lyrics categorized as country.56×56. Id. at 288. Consistent with this data, Dennis has argued that rap alone among musical genres is treated as literal expression, explaining: “courts assume that defendant-lyricists . . . do not use poetic devices or [that] the devices play a minimal role in understanding the lyrics. [Rarely] do judicial decisions explicitly acknowledge that [the] lyrics may employ metaphor, exaggeration, and other artistic devices.”57×57. Dennis, supra note 53, at 14.

Knox fits within this trend. Although it purported to examine Knox’s subjective intent, the Knox court’s two-step analysis — first examining the “content of the speech”58×58. Knox, 190 A.3d at 1158; see also id. at 1158–59. and then the “contextual factors”59×59. Id. at 1159; see also id. at 1159–61.  — failed to properly situate the song. As to the “content,” the court treated Knox’s lyrics as literal, as Dennis argues courts are apt to do, and ignored factors that belied its conclusion about the apolitical and nonartistic nature of Knox’s lyrics. The court concluded that the lyrics did not “address grievances about police-community relations or generalized animosity toward the police” or “include political, social, or academic commentary.”60×60. Id. at 1158. And yet the title of Knox’s song was a reference to N.W.A.’s “Fuck tha Police,” which famously (or notoriously) challenged racist policing.61×61. See Andrea L. Dennis, Black Contemporary Social Movements, Resource Mobilization, and Black Musical Activism, 79 Law & Contemp. Probs. 29, 47 (2016); Kubrin & Nielson, supra note 50, at 190 (explaining that the FBI was so infuriated by the song that it tried to induce N.W.A.’s record label to drop the group). The court also ignored the imagery in the music video, featuring victims of police brutality and lynching.62×62. GhettoSuperStarTV, supra note 8. Additionally, the court construed “bull horns, police sirens, and machine-gun fire” as elements designed to make the song threatening without any acknowledgement of the regular use of such sound effects in music.63×63. Knox, 190 A.3d at 1159. On the court’s analysis of the sound effects, Nielson remarked: “This is like what would happen if your grandfather tried to read rap music and understand it.” Allyson Chiu, Rapper Jailed for Song Threatening Police Tries a Free Speech Appeal — And Loses, Wash. Post (Aug. 22, 2018), https://wapo.st/2wex7Gw [https://perma.cc/PP22-USWB]. In short, the court either ignored or misconstrued the artistic and political dimensions of the song to Knox’s detriment.

Second, the court applied “contextual factors” imported from Watts and J.S. that transformed its inquiry from a purportedly subjective one into an objective one and potentially allowed for perceptions of rap as “violent” to control. In using the listeners’ reactions to Knox’s speech to discern Knox’s subjective intent,64×64. The court noted that the threats were unconditional and caused the officers to retire early, move, and obtain a security detail. Knox, 190 A.3d at 1159. the court rendered its insistence on finding specific intent meaningless. In Fourth Amendment and self-defense contexts, academics have highlighted that objective tests leave room for racial bias (that is, the notion of “black-as-criminal”) to influence judicial analysis or jurors’ perceptions.65×65. See, e.g., Cynthia Kwei Yung Lee, Race and Self-Defense: Toward a Normative Conception of Reasonableness, 81 Minn. L. Rev. 367, 383–85 (1996); Lindsey Webb, Legal Consciousness as Race Consciousness: Expansion of the Fourth Amendment Seizure Analysis Through Objective Knowledge of Police Impunity, 48 Seton Hall L. Rev. 403, 404–05, 416–18 (2018). The emphasis in Knox on the listeners’ fearful reactions to a rap song failed to cabin potential bias and vitiated the subjective prong of the inquiry.

Knox’s presumption that rap is inherently threatening was perhaps most evident in both the majority and Justice Wecht’s reliance on the personalization of the lyrics, which both opinions found set these particular rap lyrics apart. The Knox court found it significant — maybe even determinative — that the song singled out the police officers by name, indicated knowledge of when their shifts ended, and evinced a desire to kill them at home.66×66. Knox, 190 A.3d at 1159. However, personalization is a necessary but not a sufficient quality of a true threat.67×67. In Black, the Supreme Court explained that true threats reflect “a serious expression of an intent to commit an act of unlawful violence to a particular individual or group of individuals.” 538 U.S. 343, 359 (2003) (emphasis added). In other cases, courts have analyzed the detail and accuracy of the personalization to determine whether a true threat was made.68×68. For example, in Elonis v. United States, 135 S. Ct. 2001 (2015), the Court noted that the defendant included a detailed description and illustrated diagram of his ex-wife’s home in a threat, id. at 2005–06, which the Court also stressed were “accurate,” id. at 2006. See also United States v. Turner, 720 F.3d 411, 423 (2d Cir. 2013) (emphasizing the defendant’s “lengthy” discussion of killing his targets, references to a family member of one of his targets who had been killed, and detailed descriptions of how to locate his targets). Here, the court found it irrelevant that Knox’s reference to when the officers’ shifts ended might not even have been accurate.69×69. Knox, 190 A.3d at 1146. The majority and Justice Wecht’s emphasis on personalization allows it to do more work than in other true threat cases, in which personalization is merely an analytical starting point.

The Knox court’s failure to contextualize Knox’s lyrics and distinguish his intent is especially disappointing given that it initially took significant defendant-friendly steps. The court claimed to require a finding of Knox’s specific intent to threaten, a threshold few courts require.70×70. See In re Douglas D., 626 N.W.2d 725, 740 n.13 (Wis. 2001) (“[T]he vast majority of federal appellate courts and state supreme courts to have considered the issue rejected the argument that true threats require specific intent.”). The court also acknowledged the importance of contextualizing the lyrics and even cited Dennis’s article critiquing courts’ analyses of rap lyrics.71×71. Knox, 190 A.3d at 1160 (citing Dennis, supra note 53, at 23). However, the court merely acknowledged Dennis’s concerns and nonetheless conducted a literal analysis. Furthermore, the contextual factors placed a substantial emphasis on the listeners’ reactions — in this case, the fear felt by the named officers. Considering listeners’ perceptions allowed the court to condone biases toward rap despite its apparent awareness that bias could be introduced in this manner.

In order to give rap lyrics their political and artistic due and limit opportunities for bias, courts should actually apply a standard of specific intent by using “contextual factors” that emphasize the artistic process. For example: Was the song performed professionally? Was there a drafting process? Is there a larger body of work? How does this song fit into that work? How personal was the alleged threat? Was such personalization accurate and detailed? Who was the intended audience? Was the work intended to be shared at all?72×72. In another true threat case concerning rap lyrics, the prosecution rested upon a scrap of paper containing draft lyrics that was found in the defendant’s car. See People v. Oduwole, 986 N.E.2d 316, 320–21 (Ill. App. Ct. 2013). The answers to inquiries such as these could be used to create a rebuttable presumption that the song was created as art and reduce the influence of a listener’s propensity to find rap threatening.

The Pennsylvania Supreme Court’s purported reluctance to grant a categorical exemption for rap73×73. The court expressed concern that if “[Knox’s] decision to use a stage persona and couch his threatening speech as ‘gangsta rap’ categorically prevented the song from being construed as an expression of a genuine intent to inflict harm, [it] would in effect be . . . provid[ing] blanket protection for threats . . . so long as they are expressed within that musical style.” Knox, 190 A.3d at 1161. misses the point. Knox and others whose lyrics are put on trial are merely asking for their work to be treated as art rather than literal autobiography. Of course, giving rap music its artistic and political due would still bring some songs within the true threat ambit, but it would greatly limit the scope of what constitutes a true threat. True threats are meant to be only a narrow exception to First Amendment protection. To prevent the doctrine from chilling legitimate and valuable art forms, courts must not only require a showing of intent, but also be mindful to carefully contextualize subversive or unfamiliar mediums.