Gangsta' Rap and A Murder In Bakersfield
Stuart  Fischoff, Ph.D.
Dept. of Psychology
      California State University, Los Angeles
(Paper previously published in Journal of Applied Social Psychology, 1999, 29, 4, 795-805)


The present study explored the biasing effects of gangsta’ rap lyrics on subject perceptions of a murder trial defendant’s personality.  The lyrics were those actually authored by a defendant in a murder trial on which the present writer was a psychological consultant for the defense.  Results strongly indicated that the defendant was seen as more likely to have committed a murder than had he not been presented as authoring such lyrics.  Surprisingly, results also showed that the writing of such rap lyrics were more damning in terms of adjudged personality characteristics than was the actual fact of being charged with murder.

Background and Introduction
In 1995 the writer was retained by the Defense as an expert witness for a murder trial in Bakersfield, California.  This was a retrial of a murder case in which he had testified in 1992.  During the first trial, sample lyrics from a portfolio of inflammatory gangsta’ rap lyrics which were found in the home of the defendant, Offord Rollins III, an 18 year-old African-American high school student-athlete, were shown to the jury as evidence of Rollins’ alleged criminal disposition as it might relate to the murder of a Puerto Rican high school student, his former girlfriend.

The writer testified as to the poor as well as sullied evidentiary and psychological projective value of the rap lyrics.  The lyrics, while decidedly violent and misogynistic, were written for commercial and entertainment purposes.  Moreover, they were highly imitative of rap lyrics popular at the time of the murder.  Furthermore, they were co-written by Rollins and a relative.  They thus provided little value as reflecting Rollins’ state of mind or disposition to commit the murder of his former girlfriend.

The writer’s testimony notwithstanding, the first trial ended in the murder conviction of the defendant.  The lyrics were seen by the then-defense team as relevant to the jury’s decision, particularly since none of the jurors were Black, none were fans of the musical gangsta’ rap musical genre, and most of the jurors were women.  The conviction was subsequently overturned because of jury misconduct and a second trial was set.  For the retrial, a new Defense team was assembled.

Pretrial motions at the retrial pitted the Prosecution, arguing for the admission of the rap lyrics into evidence, against the Defense Attorney, H. Sala, Esq., seeking to prevent their admission.  The Defense was concerned that the lyrics would again bias the jury against the defendant, Rollins.  But proof was necessary for the presiding judge to be persuaded that the prejudicial value of the lyrics outweighed their probative value.  It was at this point that the present writer was again brought into the case.

 A review of the literature turned up a number of research studies and reports in the popular press concerning gangsta’ rap.  The popular press (e.g., Katz, 1995) reported many instances where murder defendants alleged the influence of such lyrics on their criminal behavior (e.g., Howard v Texas, 1995, Branaccio v Florida, 1995, Walker and Jackson v Wisconsin, 1994).  Interestingly, all defendants cited the inflammatory influence of the late Tupac Shakur's gangsta’ rap album 2pocalypse Now on their murderous impulses and actions.  Also of interest was the fact that all defendants were convicted of murder.  Juries across the nation were glaringly unconvinced of the contributory influence of rap music on the homicidal actions of the defendants.
Clearly, the issue of connecting a behavior or action of a person (in this instance, writing gangsta’ rap lyrics) to other personality characteristics or behaviors (in this case, to a charge of and disposition to commit murder) fell directly within the realm of person perception theory and research in general (i.e., what forces influence how we perceive and judge people) and implicit personality theory (Asch, 1946; Riemann & Angleitner, 1993) in particular.

Implicit personality theory looks at how we form judgments of people based on what we observe about them and what we infer they might do or might be because we think that certain traits tend to co-occur, e.g., fat people are happy or quiet people are thoughtful.  In the present context, the implicit personality theory connection would be that people who write ugly, violent gangsta’ rap lyrics may be disposed to murder.

Expectations of such social judgments probably go beyond the singular expectation that a person who writes violent music lyrics might also be predisposed to violent behavior.  It might also be surmised that other characteristics would acquire a negative valence if they are semantically or conceptually associated with violent tendencies.  Indeed, the early work by Asch (1946) and subsequent research by Kelley (1950), Wishner (1960), and Anderson (1981), which elaborated on Asch’s initial formulation on central and peripheral traits, suggests just such possibilities.

According to Asch (1946), central traits are those which exert a disproportionate influence on people’s overall impressions, causing them to assume the presence and the tone (e.g., positive or negative) of other traits.  Charles Winick (1979), has noted that during any phase of a criminal trial both the prosecution and the defense seek to activate impressions in the minds of jurors about the “character” of a defendant.  Often this is accomplished by such tactics as managing the physical appearance of the defendant and/or by seeking to admit into evidence information which is either favorable or unfavorable to the case of the prosecution or the defense.  This sort of impression management is designed to activate either positive or negative trait associations in the minds of jurors to engender collateral support for adjudging the defendant’s appearance of guilt or non-guilt.

In the present instance, it was expected that the admission of gangsta’ rap lyrics could have just such an effect on a jury thereby prejudicially working for the prosecution and against the defense.  Any research evidence that would demonstrate that authoring violent rap lyrics would arouse in jurors not only a connection between writing such lyrics and committing a murder but also call up other negative personality characteristics, would definitely show the consequences of admitting such lyrics into evidence when weighing their probative against their prejudicial value.  More specifically, such research evidence might highlight the strong possibility that when jurors are exposed to such defendant image-impairing lyrics, they might become more disposed to and confident in a guilty verdict what with the added weight of the negative personality trait associations conjured up by such inflammatory lyrics.

Such expectations are consistent with the work of Asch (1946), and are also consistent with expectations deriving Heider (1958) and his early work on balance theory.  Heider argued that people seek cognitive consistency in their impressions of trait constellations in the area of social judgment and person perception.  In other words, bad people do bad things and possess other, bad traits.  Previous research on the jury decision making process (cf., Fischoff, 1979), supports related balance theory predictions.

Unfortunately, the academic research extant at the time of the literature search fell short of the explicit data demanded to meet forensic exigencies.  Some research focused on listener reactions to gangsta' rap lyrics.  Bleich, Zillmann, and Weaver (1991) and Hansen-Hall (1995), for example, demonstrated the biasing effects of listening to rap lyrics in terms of arousing ugly passions and attitudes.  Presumably the same or similar research was at the core of the defenses’ arguments in the murder trials mentioned above.  Research by Tapper, Thorson and Black (1994) established that, compared with other music genres such as Soul, Country, Heavy Metal, Pop, Classic Rock, the music videos of gangsta’ rap contain twice the lyrical and visual violence.

Thus, while the above literature addressed the topic of gangsta’ rap, no studies were reported which pertained to how real or mock juries might regard a murder defendant who had actually written violent, misogynistic lyrics as compared with a murder defendant who had simply listened to such lyrics.  Nor in fact, was there any research on how a gangsta’ rap lyricist might be perceived or judged by Participants in terms of personality traits.  The explicit and necessary connection of trial particulars-related research results and research paradigms to the implicit personality issue was, consequently, absent.
In order to fill the research gap and investigate the questions of how possible jurors might perceive or evaluate murder trial-relevant personality traits of a gangsta’ rap lyricist and how such perceptions might, inferentially, affect judgments of a defendant in a murder trial, with the permission of the defense attorney and defendant, the following study was undertaken.



Students at California State University, Los Angeles comprised the participant sample of 134.  Ages ranged from 18-56 with a mean age of 27.6 years.  There were 56 males and 78 females in the sample.  Selection and assignment of participants assured that Asians, Whites, Blacks and Hispanics were equally represented across conditions.


Each participant was randomly assigned to one of four conditions.  Each condition provided variable but factually descriptive biographic information about a Target Male (Offord Rollins, the actual defendant in the murder case), hereafter referred to as the Target Male.  The participants, however, were not informed of the factuality of this information or that the hypothetical details were related to an ongoing criminal trial.  They were simply told to respond to the information provided about the Target Male on a series of nine bi-polar adjective scales which were selected because they were considered relevant to the particulars of both the murder case and the social and personality characteristics of the defendant.  After the participants completed the task, they were asked if any of them were familiar with the case.  None were.  This was to be expected insofar as the case was not publicized in the Los Angeles area media and Bakersfield is over one hundred miles north of Los Angeles.


Verbatim Descriptions and Instructions Provided Participants in Each of the Four Conditions

 Condition 1: No Murder-No Lyrics.  An 18 year old African American male high school senior resides in the Southern California region.  He is a state champion in track, has a good academic record and is planning on attending college on an athletic scholarship.  He makes extra money by singing at local parties.  On the following page, give your impressions of the young man by using the rating scales provided.

Condition 2: Murder-No Lyrics.  An 18 year old African American male high school senior resides in the Southern California region.  He is a state champion in track, has a good academic record and is planning on attending college on an athletic scholarship.  He makes extra money by singing at local parties.  He is on trial accused of murdering a former girlfriend who was still in love with him, but has repeatedly declared that he is innocent of the charges.  On the following page, give your impressions of the young man by using the rating scales provided.

 Condition 3: No Murder-Lyrics.  Below are a set of rap lyrics written by an 18 year old African American male high school senior residing in the Southern California region. He is a state champion in track, has a good academic record and is planning on attending college on an athletic scholarship. He makes extra money by singing at local parties.  Read the lyrics and then, on the following page, give your impressions about the male by using the rating scales provided.

 Condition 4: Murder-Lyrics.  Below are a set of rap lyrics written by an 18 year old African American male high school senior residing in the Southern California region. He is a state champion in track, has a good academic record and is planning on attending college on an athletic scholarship. He makes extra money by singing at local parties. He is on trial accused of murdering a former girlfriend who was still in love with him, but has repeatedly declared that he is innocent of the charges  Read the lyrics and then, on the following page, give your impressions about the male by using the rating scales provided.

The set of lyrics presented to participants in Conditions 3 and 4 are displayed below.  They are part of a portfolio of such lyrics which Rollins co-authored and which the Prosecution was seeking to present into evidence at the trial.  The lyrics were presented to participants in their original form, misspellings and all, so as to not undermine their authenticity and whatever associated impressions such misspellings and vernacular might convey.

 "Id die before
"my dick starts to fizz
"pulled it out
"and my head smelled like fish
"rush to the shower
"to wash my dick
"Let me go, Let me go
"Bitch let me go
"She wouldn't let me go
"So I slaped the ho
"don't get mad
"You fruit cocktail
"See my ryhmes
"Now you happy
"like a fag in jail
"sayin my name wrong
"you trick silly rabbit
"come in my face again
"I'm gonta grab it
So watch your
"chains and Nugget
"cause with the Steel
"in my hand I'm ruggit
"put the gaurd up for your gold teeth
"you littel fink
"talk one more line
"then I'm a sluggit"

The nine, murder trial-relevant, bi-polar adjective scales to which the participants responded were: caring-uncaring, selfish-unselfish, gentle-rough, likable-unlikable, conceited-modest, truthful-untruthful, sexually nonaggressive-sexually aggressive, capable of murder-not capable of murder, not a gang member-a gang member.  They were presented in a 6-point Semantic Differential scale format and weighted during analysis such that the higher the score the more negative the evaluation of the described male.


ANOVAS and t-tests were conducted on individual trait scale scores and on the total score when summing across all trait scales.  Results revealed consistent and significant main effects for Murder (Murder, No Murder), main effects for Lyrics (Lyrics, No Lyrics) and interaction effects.
 Total Score Outcome
As Table 1 and the accompanying Figure 1 indicate, in terms of total score, results were dramatic in their demonstration that participants viewing violent, misogynist rap lyric samples judged the Target Male far more negatively than when he was not associated with such lyrics, F (3, 130) = 132.83, p < .001, Mean Rap Lyrics = 41.23, Mean No Rap Lyrics = 27.56.  Furthermore, a Target Male rap lyricist accused of murder was viewed significantly more negatively than the Target Male non-lyricist accused of the same murder, F (3, 130) = 17.21, p <.001, Mean Lyrics-Murder = 42.65, Mean  No Lyrics-Murder = 31.06.

Post-hoc analysis of the interaction between Murder and Lyrics using a Tukey HSD Multiple Comparison test also indicate that participants were significantly inclined to more negatively evaluate a gangsta' rap lyricist not accused of murder than a nonlyricist accused of murder, p < .001, Mean = 39.8 and Mean = 31.06 for Lyrics-No Murder and No Lyrics- Murder respectively.

Table 1
Condition Means for Individual Trait Scores and Total Scores


Figure 1
Total Scores, By Condition, On Personality Judgment Scales

Individual Scale Item Outcomes

As mean scores in Table 1 reveal similar results obtained for trait items viewed individually.  The Target Male rap lyricist received more negative evaluations than the Target Male non-lyricist on all trait dimensions (9 out of 9), with the differences being statistically significant on 8 out of 9 trait items.  However, and unexpectedly, the results were less dramatic as regards whether or not the Target Male was on trial for murder.  In this case, results were more negative on 8 out of 9 trait items and significantly different on only 6 out of the 9 dimensions.  Clearly, participants were more put off by the rap lyrics than by the murder charges.

Returning to the issue of total score, the interaction between Lyrics and Murder for the total score approached but did not achieve statistical significance, F (3, 130) = 3.04, p < .08.  The post-hoc analysis revealed, however, that, while there was a predictably significant difference between No Murder - No Rap (Mean = 24.07) and Murder - No Rap (Mean = 31.06) conditions, p < .001, the difference between Murder - Rap (Mean = 42.65) and No Murder - Rap (Mean = 39.80) conditions was not significant, p < .30.  This reinforces the conclusion that authoring gangsta’ rap lyrics seemed to be at least equally potent with murder charges in impact on judgments.  As Table 1 also reveals, similar results obtained for trait items viewed individually.

Analysis of data by Gender revealed significant differences on only one of the variables, “conceit”.  Males tended to evaluate the defendant as more conceited than females, t (132) = 3.68, p < .001, Mean = 4.61 for Males and Mean = 3.95 for Females.  Males, perhaps, viewed the Target Male from a more competitive vantage point than females.


Study results clearly indicate that showing participants the rap lyrics exerted a significant prejudicial impact on the evaluation of a person and particularly so when the person is accused of murder.  From the point of view of implicit personality theory, this would indicate that, in the minds of the participants in the present experiment, writing such lyrics invited strong associations with inferences about other negative traits.  This would suggest, perhaps, that nice males don’t write ugly lyrics and that males who do are emphatically not nice.  It would also suggest that authoring gangsta’ rap lyrics vies with being charged with murder in terms of the impact of central trait properties in the person perception process.

That exposure to the lyrics evoked a negative reaction in participants that was more intense than the reaction to being told that the young man was on trial for murder, was unexpected.  But why would writing gangsta’ rap lyrics evoke more negative reactions than being accused of murder?  Results may be explained by the fact that the facticity of the lyrics and their authorship by the person being judged were not in contention.  The Target Male (Rollins) wrote them, the participants viewed them and they were offended by them.  The charge of murder, on the other hand, was just that, a charge.  The outcome of the trial was unknown and therefore uncertain.  He may have committed the murder, but he may not have.  Yet, he definitely wrote the lyrics.  Perhaps this certainty with regard to the latter and uncertainty with regard to the former contributed to the more negative evaluation of lyric writing than did that of the murder charges.

The unambiguous results of the personality judgments of the rap lyricist are chilling in their implications.  Creative expression in any art form, music lyrics, fiction writing, screenplays, even actors portraying roles, may be seen by the public as an authentic expression of personality.  On the positive side, John Wayne, as a glaring example, who starred in numerous war movies, was venerated by many as a military hero.  This was so even though Wayne never actually served in the military.  On the negative side, it is well-known in Hollywood that William Smith, the actor who portrayed the villain, “Falconetti,” in the 1977 mini-series Rich Man, Poor Man, and whose part required that he kill “Tom Jordache” (the young Nick Nolte), suffered insults, assaults, and occasional battery in person and by phone for years after the show aired.  He couldn’t shake the “bad guy” public image.  It even affected his choice of roles in film and television.  Clearly, in Smith’s case, no good deed (good acting) goes unpunished.  Similarly, it took Ernest Borgnine years to shake the villainous image of “Fatso Judson” in the 1953 film From Here To Eternity.  He played the part extremely well.  And people did not let him forget that he killed Frank Sinatra’s character in the movie.  For Smith and Borgnine, and, yes, John Wayne, in the eyes of the public, you are what you portray!
But do people really believe that horrormeisters such as novelist and screenwriter Stephen King or Nightmare On Elm Street writer-director Wes Craven are disposed to violence and murder because they make a living presenting fictionalized and fantastic accounts of such themes for public consumption and entertainment?

Perhaps not.  But what if such artists were accused of and stood trial for murder?  Might the public and the jurors infer that there is a connection between what one writes about as fiction or for entertainment purposes and what one is disposed to do in reality?  Based on the present research results, the outcome of the first trial and the desire by the Prosecution to get the gangsta’ rap lyrics into evidence in the retrial, it seems that people may indeed be inclined to identify an artist with his/her artistic product.

A final, cautionary note:  Further research on the subject of identifying an artist-defendant at a criminal trial with his or her artistic work is warranted.  It is conceivable that judging an artist-defendant on similar scales, as part of a social judgment task stemming from mock jury group deliberations, may produce results different from those obtained in the present study where participants were not instructed to think of themselves as jurors or to engage in group deliberations.  Moreover, varying the race or ethnicity of the defendant may also be found to serve as a moderating influence on such social judgments if it were the case that, e.g., a White defendant who authored gangsta’ rap lyrics might be adjudged differently than a Black defendant.  This possibility would suggest incorporating such a demographic consideration into subsequent research designs.

Details of the study were presented at the pre-trial hearings of the Rollins retrial.  Based on the present writer’s testimony and the results of the research, the majority of the lyrics were excluded from evidence during retrial and the most inflammatory lyrics were completely excised.  The retrial ended in a deadlocked jury.  On September 16th, 1996, the District Attorney of the Kern County decided not to retry the defendant for a third time.  As of this writing, the former defendant is a member of the varsity track team of a local junior college.


Anderson, N.H. (1981).  Foundations of information integration theory.  New York: Academic Press.

Asch, S. E. (1946). Forming impressions of personality.  Journal of Abnormal and Social Psychology, 41, 258-290.

Bleich, S., Zillmann, D., and Weaver, J. (1991).  Enjoyment and consumption of defiant rock music as a function of adolescent rebelliousness. Journal of Broadcasting & Electronic Media, 35, 351-366.

Fischoff, S. (1979). “Recipe for a jury" revisited: A balance theory interpretation. Journal of Applied Social Psychology, 9 (4), 335-349.

Hansen, Christine-Hall (1995). Predicting cognitive and behavioral effects of gangsta’ rap.  Basic and Applied Social Psychology, Feb, Vol 16 (1-2), 43-52.

Heider, F. (1958). The psychology of interpersonal relations.  New York: Wiley.

Katz, J. (1995, August 5). Rap furor: New evil or old story?  The Los Angeles Times, pp. A1, A23.

Kelley, H. H. (1950).  The warm-cold variable in first impressions of persons.  Journal of Personality, 18, 431-439.

Riemann, R., & Angleitner, A. (1993).  Inferring interpersonal traits from behavior: Act prototypicality versus conceptual similarity of trait concepts.  Journal of Personality and Social Psychology, 64, 356-364.

Tapper, J., Thorson, E., and Black, D. (1994).  Variations in music videos as a function of their musical genre.  Journal of Broadcasting & Electronic Media, 38, 1, 103-114.

Winick, C. (1979).  The psychology of the courtroom.  In H. Toch (Ed.), Psychology of crime and criminal justice (pp. 68-105).  New York: Holt, Rinehart and Winston.

Wishner, J. (1960).  Reanalysis of “impressions of personality,”  Psychological Review, 67, 96-112.

Return to home page