Journal of Social Issues, VoL 53, No. 3, 1997. pp. 563-581

Cognitive-Affective Processing System Analysis ofReactions to the O. J. Simpson Criminal IVial Verdict

Rodotfo Mendoza-Denton**Columbia University

Ozlem N. AydukColumbia University

Yuichi ShodaUniversity of Washington

Walter Mischel*Columbia University

Reactions to the O. J. Simpson verdict were analyzed using the Cognitive-AffectiveProcessing System (CAPS) model. Content analyses of participants' open-endedreactions to the verdict revealed that differences in the accessibility of cognitive-affective units arul their subsequent activation pathways characterized respon-dents ' reactions, but participants' race appeared to have no direct effect. Theresults were used to construct cognitive-affective domain maps that underlayelated, dismayed, and ambivalent reactions. By promoting a deeper understandingand appreciation of reactions to the verdict, we believe the domain maps facilitateovercoming the widespread tendency to attribute the cause of divergent reactions toan individual's race. The results have implications for the understanding of cul-tural differences.

•NIMH grant MH39349 to Walter Mischel supported the preparation of this article.**The authors would like to thank Liz Ekehlad, Alessandra Testa, and Anita Sethi for their help in

preparing the manuscripLCorrespondence regarding this article should be addressed to Rodolfo Mendoza-Denton, Columbia

University, Department of Psychology, 406 Schermethom Hall, New York, NY 10027, Electronic mailshould be sent via Internet to rodolfo@psych.columbia.edu.

563

©l997TlitSociayfofll>e Psychological Study of Social Issues

564 Mendoza-Oenton, Aydtik, Shoda, and Miscbel

The O. J, Simpson criminal trial verdict has perhaps been the strongestreminder in recent history that Blacks and Whites in America live in differentworlds. Images of cheering Black faces juxtaposed with those of dismayed Whitefaces across front pages of newspapers and magazine covers illustrated how peoplewho live in the same world can understand it and experience it differendy. The samecase, the same evidence, and the same trial led to two virtually opposite reacdons tothe same verdict.

Part of the condnuing fascination with the Simpson case lies in the almost irrec-oncilable nature of the two predominant viewpoints about the criminal trial verdict:one claiming that Simpson's acquittal was a travesty of jusdce, the other that any-thing but his acquittal would have been a travesty of justice. Moreover, the inabilityof either side to see the other's p)oint of view has provided little common ground fordiscussion or reconciliadon. In this article, we aim to provide some common groundby exploring reactions to the verdict not as the thoughtless, irradonal responses por-trayed by opposing sides, but rather as a result of the thoughts and feelings that thecase evoked for different people. We hope that an understanding of the processesthat brought about the reacdons to the verdict will promote a greater understandingof the world of the "other,"

On October 3, 1995, Simpson was acquitted of the murders of Nicole BrownSimpson and Ronald Goldman, Reacdons to the verdict came swiftly, and they wereemodonally charged (Margolick, 1995). "It's about racism," explained a Black sup-porter of the verdict to the New York Ttmes on the day of Simpson's acquittal, "today,racism took a solid blow." Those who disagreed expressed equally strong opinions,"With all the evidence against him, his history of abusing his wife," lamented aWhite man, "it's amazing," As easy as it may be to classify agreement and eladon asthe "Black" response and disagreement and dismay as the "White" response, how-ever, such generalizadons are unlikely to be useful in understanding people's reac-dons. An adequate explanadon of this phenomenon requires an account of theprocess that brought about people's reacdons to the verdict—one that mediated therelationship between skin color and reacdon. Understanding this process allows usto take our explanadon out of the Black vs. White dichotomy and enables us to makesense of those Whites who may have cheered and those Blacks who may havegrieved, as well as those who were ambivalent.

We propose a dynamic model, based on Mischel and Shoda's (1995) Cogni-dve-Affecdve Systems Theory of Personality, to account for the many reacdons tothe verdict within a single, unifying framework. By integrating cumuladve findingsand models from personality and social psychology in general and research in socialcognidon in pardcular (e.g,, Bandura, 1986; Bargh, 1994; Block, 1995; Cantor &Kihlstrom, 1987; Carver & Scheier, 1981; Cervone, 1991; Kihistrom, 1990;Mischel & Shoda, 1994; Pervin, 1994; Ross & Nisbett, 1991), the model allows usto understand reacdons to sociopwlidcal informadon in terms of the cognidons andaffects such informadon acdvates within individuals. Guided by this model, we

CAPS Analysis of Reactions to the Simpson Case 565

present data fTom a study in which pardcipants' reacdons to die verdict were contentanalyzed for the hypothesized cognidons and affects.

Representing "Worlds": The Cognitive-Affective Processing System

Over the last decade, models of human informadon processing have shiftedaway from serial, centralized processing models based on the architecture of tradi-donal digital computers to more sophisdcated connecdonist models in which largeamounts of informadon are processed simultaneously within a parallel, distributedsystem (Read & Miller, in press). Although there are many specific variadonswithin this direcdon, current models of cognition share the premise that informadonprocessing "units" (e.g,, mental representadons) are organized within a network ofinterreladonships that guides and constrains their acdvadon (Anderson, 1988;Bower, 1978; Kunda &Thagard, 1996; Read & Miller, 1993; Rumelhart& McClel-land, 1986), Applicadons of these information processing models to higher levelphenomena, such as attitude change (Sp>ellman, Ullman, & Holyoak, 1993),explanatory coherence (Thagard, 1989), and disposidonal inference (Read &Miller, 1993) suggest a new, powerful way of understanding how human beingseffecdvely process the large amounts of informadon with which they are constantlyconfronted,

Mischei and Shoda's (1995) Cognitive-Affective Processing System (CAPS)Theory is an approach to personality that integrates many insights from cognidveand social theories of social informadon processing (e,g,, Anderson, 1983; Ban-dura, 1986; Cantor & Kihlstrom, 1987; Carver & Scheier, 1981; Dodge, 1986;Hinton, McClelland, & Rumelhart, 1986; Mischel, 1973,1990) within a compre-hensive theoretical framework. As a connecdonist model, it proposes that behav-ior is mediated by a set of cognitive-affective units (CAUs) organized within astable activadon network that reflects the individual's learning history. Table 1summarizes the proposed CAUs, which have been idendfied as playing a role insocial behavior generadon (Alston, 1975; Read, Jones, & Miller, 1990; Pervin,1989, 1994),

The model postulates that of all the beliefs, goals, values, encodings, andfeelings one can potendally experience at any given dme, only those acdvated caninfluence subsequent behavior. Although some cognidons are acdvated onlywhen the immediate situation brings them to mind, recent research indicates thatthere may also be individual differences in the CAUs' accessibility—that is, in thedegree of readiness with which constructs in memory can become activated(Bruner, 1957; Higgins, 1990, 1989; Higgins & King, 1981; Higgins & Chaires,1980),

To understand the pjerson fully, however, it is important to have an idea of howthe person's mediating units relate to one another. The model conceptualizes theperson not just as a list of CAUs; instead, it proposes that die units are organized into

566 Mendoza-Denton, Ayduk, Sboda, and Miscfael

Ihbte 1, Types of Cognitive-Affective Units in the Personality Mediating System

1. ENCODINGS: Categories (constructs) for the self, people, events and situations (extemal andinternal).

2. EXPECTATIONS AND BELIEFS: About the social world, about outcomes for behavior in particu-lar situations, about one's self-efiicacy.

3. AFFECTS: Feelings, emotions, and affective responses (including physiological reaxmons).

4. GOALS: Desirable outcomes and affective states, aversive outcomes and affective states, and goalsand life projects.

5. COMPETENCIES AND SELF-REGULATORY PLANS: Potential behaviors and scripts that onecan do as well as plans and strategies for oi;ganizing action and for affecting outcomes and one's ownbehavior and internal states.

Note: From "A Cogtutive-Affective System Theory of Personality: Reconceptualizing Situations, Dis-positions, Dynamics, and Invariances in Personality Stnicttires," by W. Mischel and Y. Shoda, 199S,Psychological Review, 102, p. 253. Copyright 1995 by the American Psychological Association. Re-printed with permission.

a unique network of interconnecdons that funcdon as an organized whole. Posidve(excitatory) connecdons to a CAU increase its acdvadon level, whereas negadve(inhibitory) connecdons to it decrease its acdvadon level. Figure 1 shows a highlysimplified, schemadzed version of a CAPS,

TTie figure illustrates a CAPS network's distinguishing characterisdcs, Ap>erson (the large outer circle in the middle of the diagram) is characterized by apardcular subset of CAUs (the nodes within the larger circle) accessible to him orher as well as by the network of acdvadon (solid arrows) and inhibidon (brokenarrows) among those units. When a person encounters a pardcular situadon, theCAPS is sensidve to pardcular features of that situation, which become encodedand acdvate situadon-relevant CAUs within the system. These units, in turn, makeother cognidons and affects accessible while inhibidng others.

As the individual grows, leams, and gains life exp)erience, an increasingly richand complex network of CAUs develops. Whereas some CAUs are acquiredthrough the individual's unique experience (Mischel, 1973, 1990; Mischel &Shoda, 1995), others seem to be shared and transmitted among members of culturalgroups (Geertz, 1973; Gordon, 1982; Kluckholn, 1965; Obeyeskere, 1981; Tylor,1871), Thus, life experiences that members of a group share are likely to generate aculturally shared CAPS network embedded within an individual's processingsystem. If features of a situation acdvate this culturally shared network, individualsmay generate similar reacdons to that situadon. We propose that the Simpson trialwas one such situadon.

CAPS Analysis <rf Reactions to the Simpson Case 567

Cognitive-affective personality system (CAPS)

. Encoding Behavior generatkmprocASS process

• Interactions among mediators ---

Fig, 1. Schematic illustration of the Cognitive-Affective Processing System. From "A Cognitive-Affective System Theory of Personality: Reconcepmalizing Situations, Dispositions, Dynamics, and In-variance in Personality Structtues," by W. Mischel and Y. Shoda, 1995, Psychological Review, 102. p.254. CopjTight 1995 by the American Psychological A.ssociation. Adapted with pertnission of theauthors.

A CAPS Model of Reactions to the Simpson Verdict

An informal review of repwrts in the media following the verdict suggested thatMark Fuhrman's racist atdtudes, as well as a percepdon that Simpson was beingtreated unfairly because he was Black, led diose who agreed with the verdict tobelieve it was correct. Among those who disagreed, on the other hand, dismay anddisbelief were often accompanied by mendon of the preponderance of the evidenceagainst Simpson as well as Simpson's past history of spousal abuse.

We proposed that differences in die nature of and reladons between CAUswithin the processing system could account for the divergent reacdons surroundingthe Simpson verdict. We did not exp>ect that race per se determined one's opinion tothe verdict. Instead, we hypothesized that the role of race was mediated through thethoughts and affects that the case acdvated for different individuals. Specifically, weexpected CAUs related to discriminadon against Blacks to form part of the cultur-ally shared CAPS network of a Black pjerson, making racism by the p)olice and the

568 Mendoza-Denton, Ayduii, Shoda, and Miscfael

courts more chronically accessible and highly affect letden. In contrast, we expectedthat for a White person, discriminadon was more likely to be a "cool," abstract cog-nidon (see Mischel, Shoda & Rodriguez, 1989) that might more easily be inter-preted as having nothing to do with the case. As such, we postulated that amongthose for whom racism was not a highly affecdve, "hot" issue, cognidons about wifeabuse and the "mountain of evidence" became the focus of the Simpson case. Inaddidon, we expected that among those for whom abuse, evidence, and racismissues were equally accessible, the reaction to the verdict would fall somewherebetween agreement and disagreement—namely, ambivalence toward the verdict.

To test our hypotheses, we conducted a study in which pardcipants were askedto describe their reacdons to the verdict. We chose to have pardcipants write theirresponses in a free-form essay format because that would allow us to detect thekinds of cognidons and affects that pardcipants spontaneously generated as well asto examine the reladons among CAUs the people would mention.

Method

Participants

One hundred and thirteen people (49 females and 64 males) who were either tak-ing classes or working at Columbia University during the summer of 1996 pardci-pated in the study, 29.2% of the pardcipants idendfied themselves as Caucasians,33.6% as African Americans, 9,8% as Ladnos, and 22.1% as Asians and AsianAmericans. Pardcipants from other backgrounds accounted for 5,3% of our sample.The mean age was 21.7 (range: 14 to 49). There were no differences in the mean age ofpardcipants among different racial/ethnic groups. Educadonal levels were 36.3%high school students, 37,2% college students and 24.8% graduate students, with 1,7%not currently students, PardcipanLs were paid $5 for their participadon.

Materials

Eighteen movie and sound clips related to the Simpson trial were downloadedfrom the CfW (http://www,cnn,com) and Court TV (http://www,courttv,com)World Wide Web sites onto a Power Macintosh 7500, The clips were specificallyselected to recreate as vividly as pwssible the trial's most dramadc asp)ects in roughtempx)ral sequence, from the Bronco chase to the presentadon of the evidence to theverdict and the families' reacdons to it (see Appendix A),

Procedure

Pardcipants were recruited by way of flyers posted around the campus. Eachpardcipant first signed a consent form and was assured that his or her resp>onses

CAPS Analysis of Reactions to tlie Simpson Case 569

would be kept strictly confidendal and anonymous, Pardcipants were brought into acomfortable, softly lit room and were seated in front of a Macintosh computer. Theythen viewed each of the 18 clips from the trial in succession; this procedure tookapproximately five minutes. Having watched all the clips, participants were led toanother room and were given a pen and a blank wridng pad. They were instructed towrite down their reacdons to the verdict, with sp)ecific instrucdons to specifywhether they agreed, disagreed, or were somehow ambivalent about the verdict.When they were done with their essays, they were debriefed and paid.

Coding Protocol and Intercoder Reliability

To idendfy the cognidve-affective units that pardcipants spontaneouslygenerated, their essays were subjected to content analysis following the procedureLampert and Ervin-Tripp (1993) outhned. Twenty-two representadve cognidve-affecdve units as well as seven emotional reactions specific to the verdict were iden-dfied (see Table 2), Two indep)endent judges coded each essay according to whetherit mendoned each of the cognidve-affecdve units and/or emodonal reacdons. If aunit was not represented in an essay, that unit was coded 0, If it was mendoned onlyonce, it was coded 1. If it was mendoned more than once within the same essay, itwas coded 2. Each essay was also coded for agreement, disagreement, or ambiva-lence toward the verdict.

Intercoder reliability results revealed a high degree of agreement between thetwo coders (median Cohen's kappa = ,84, range = .49 to 1,0). A third coder arbi-trated any conflicdng codings of the essays to generate the finai coding, which wasused in all subsequent analyses.

Operationalization of Accessibility and Activation Pathways

Accessibility of thoughts and affects was conceptualized in the present analy-sis as the frequency with which each CAU was mentioned among those whoagreed with (A = 45), disagreed with (N = 46) or felt ambivalent about the verdict(A' = 22), Only those CAUs that were differendally accessible to one group overthe others, as revealed by chi-squares, were thought of as belonging to that group.The links between these accessible constructs were identified by way of correla-tions computed among all the constructs coded for, CAPS theory operates on theassumpdon that individual differences exist in accessibility of constructs, as wellas differences in the networks of acdvadon. In diis ardcle, we concentrate primar-ily on tesdng the first of these assumpdons, while holding the second assumptionconstant That is, as a first approximadon, we assumed that relevant pathways aresocietally shared given the larger sociopxjlidcal context. Therefore, correiadonswere computed using all pardcipants, A positive correlation between two itemswas interpreted as an excitatory link, meaning that the acdvadon of one construct

570 Mendoza-Denton, Aydufc, Sboda, and Mischel

led to the acdvation of another, A negadve correladon, by contrast, was inter-preted as an inhibitory link; that is, the acdvadon of one construct led to the inhibi-don of another.

Construction of Cognitive-Affective Domain Maps

Based on both the accessibility (frequency) and the acdvadon pathway (inter-correladons) analyses, we modeled the CAPS that generated different emodonalreacdons to the verdict. Several rules were followed in construcdng the domainmaps. If a thought was significandy more accessible to one group than another, itwas grotmds for drawing it into the former's domain map. The darkness of eachCAU node depicted the strength of accessibility of each CAU (i.e., the frequencywith which it was mendoned within each group): the darker the node, the moreaccessible the construct. Inhibited nodes, by contrast, were depicted with dashedovals, (Figures 2 through 4 present several of the maps constructed.)

In represendng the acdvadon pathways between the units, only those links sig-nificant below the ,01 level in the correladon analyses were drawn into the domainmaps. Solid arrows represented ptosidve correiadons between thoughts (i.e., excita-tory links), whereas broken arrows represented negadve correiadons (i,e., inhibi-tory links). Even though the inhibited thoughts themselves had acdvadon links toother cognidons, we chose not to draw them in, assuming that an inacdve node doesnot influence other nodes in the system (see Shoda and Mischel, 1996, for a detaileddiscussion).

TTie only excepdon to the rule above involved the links to the nodes "the verdictis wrong," "the verdict is correct," and "I don't know whether to agree or disagree,"for which a stricter criterion was applied. That is, only those links diat significandypredicted agreement, disagreement, or ambivalence when controlling for all otherfactors (all CAUs accessible to the group) were drawn in as links to any of thosenodes. Similarly, only those thoughts that significantly predicted the emodonalreacdons to the verdict when controlling for all other variables (all CAUs as well asthe opinion itself) were drawn as behavioral output links.

The input to the processing system was operadonalized as the video andsound clips presented to the participants. Those clips that represented similaraspects of the trial were grouped together as a "feature"; for example, the twoCNN computer animadons of the actual crime scene were grouped together as the"crime scene" feature. Similarly, the three audio clips in which the defense attor-neys spoke were grouped together as "defense arguments," Because the study'sdesign did not allow us to extrapolate the specific links between situadonal fea-tures and the cognidons and affects they directly acdvated, we have assumedthose acdvation links based on the conceptual similarity between input featuresand the CAUs,

CAPS Analysis of Reactions to tlie Simpson Case - 571

Results

Accessibility of CAUs

Chi-square analyses conducted on the frequency of CAUs across differentgroups revealed significant accessibility differences among those who agreed withthe verdict (the elated), those who disagreed (the dismayed), and those who couldnot make up their minds (the ambivalent).

tbble 2. Percentage of i>articipants for Whom Each Cognitive-Affective Unit and EmotionalReaction Was Accessible, as a Function of Their Opinion about the Verdict

Cognitive-Affective Units

Simpson was aggressive, abusiveSimpson behaved as though guiltyCourt system long biased against BlacksSimpson assumed guilty because BlackJury was brave, correctDefense devious, played race cardRacism no excuse for murderSimpson assutned innocent because famousIf he did it, he'll answer to GodProsecution was goodDefense was greatHe knows about the murdersMoney buys you freedomThere is a mountain of evidenceAbuse doesn't mean mtirderEvidence is questionableMark FuhrmanyLAPD is racistNicole Brown Simpson was no saintJtiry was stupidThe justice system sucksVictims and their families sufferedJustice was doneATictory for Black peopleEMOTIONAL REACTIONSAngryConfusedHappyRelievedSadShockedUpset

Disagree(dismayed)

57.48a

4,OaOa

26,46a28a

2a

2a9a

2a30,74«

2aOa

22a

Oa20a

35a

I3..b0.

n.Oa2a

Oa

9a

Ha30a

Agree(elated)

24b

4b22b27b

I3b

4b2b4b

13a

Oa22a20b

9b

9b18b76b44b

I6b

Ob

4b9b

27b

7.4.

27b20b

Ob4.4t

Ambivalent

36aJ,18b32b

9b

Oa,b

9^b27.14a,b

Oa5a5.Oa

18a,b

32c5a,b

23c23a.b

5a,b

Ob18a,b

32.9b

14a

Oa18b

5a,b

18a14.18»,b

Note: Percentages with different subscripts differ significantly atp< ,05 by the chi-square significantdifference te.st; chi-square comparisons of cells with small N may be unreliable.

572 ' Mendoza-Denton, Ayduk, Sboda, and Mischel

As Table 2 shows, the elated accessed the cognidons "Court system long biasedagainst Blacks," "Simpson assumed guilty because Black," and "Fuhrman/LAPD isracist" significandy more frequendy than the dismayed. By contrast, the dismayedaccessed "Racism no excuse for murder" and "Defense played race card" signifi-candy more frequendy than the elated. With respect to domesdc abuse, "Simpsonwas aggressive, abusive" was more accessible to the dismayed than the elated,whereas "Abuse doesn't mean murder" was more accessible to the elated than thedismayed. Furthermore, the dismayed accessed "There is a mountain of evidence"more frequendy than the elated. "The evidence is quesdonable" was highly acces-sible to the elated, yet nonexistent among the dismayed. The elated greeted the ver-dict with a combinadon of happiness and relief, whereas the dismayed were mostlyupset.

The ambivalent, unlike the elated, recognized that die evidence overwhelm-ingly pointed at Simpson and that the vicdms and families suffered. Unlike the dis-mayed, however, they recognized that the court system has long been biased againstBlacks, Moreover, the ambivalent felt that the evidence was quesdonable less oftenthan the elated, but more often than the dismayed. The apparent conflict between theaccessible cognidons of the ambivalent mirrored itself in their emotional reacdonsto the verdict. Happiness was more conunon for the ambivalent than the dismayed,whereas sadness was more accessible to them than it was to the elated.

Activation Pathways

Elated Activation Network. All the CAUs accessible to the elated were posi-tively correlated with agreement with the verdict. However, to establish whichCAUs played an independent role in activadng agreement, we regressed agreementon the accessible cognidons within the system (F(9,103) = 14,3, R'= ,56,/7< ,001).Results revealed that, when controlling for the effects of other CAUs, recognidonthat the evidence was quesdonable predicted agreement with the verdict (yS = ,31,p < .001), In part because Simpson was perceived as prejudged (fi = .2,p< .05),Addidonally, the idea that Simpson may know something about the murders, even ifhe is not gtiilty, predicted agreement (j8 = ,37, p < ,003).

To determine which CAUs played an independent role in bringing about eladonin response to the verdict, we regressed agreement together with all the other acces-sible constructs in the processing system both on relief and on happiness (F (10,102) = 3.85, R^ = ,27, p < ,001, and F(10, 102) = 3.62, R^ = ,26, p < .001, respec-dvely). Results indicated that the elated felt relieved because Simpson wasacquitted despite his being a Black man (fi = .2,p< ,005), On the other hand, per-ceiving the verdict as fulfillment of jusdce and a victory for Blacks predicted happi-ness in response to the verdict (/? = ,28, p < ,005),

CAPS Analysis of Reactions to the Simpson Case 573

Dismayed Activation Network. All the CAUs accessible to the dismayed hadpositive correiadons with disagreement. To establish which CAUs had an independ-ent effect on disagreement, disagreement was regressed on the accessible cogni-dons in the system (F(9, 103) = 14.8, P^ - .56, p < ,001). Results indicated diatthinking that the evidence was overwhehning (fi = ,24, p < ,(X)1), that Simpson'sbehavior was suspicious (fi = .l5,p< ,04), that his celebrity status blinded people tohis guilt (fi = .2\,p< ,03), and that the jury was stupid (fi = .23,p < .03) all predicteddisagreement with the verdict. Subsequent regression analyses on being upsetincluding disagreement as an independent variable (F (10, 102) = 2.29, R'= , 18, p< ,02) indicated that this reacdon to the verdict was solely predicted by disagree-ment (fi = .24, p < ,02), TTie model was not significant for sadness.

Ambivalent Activation Network. To establish which CAUs had an independenteffect on ambivalence, ambivalence was regressed on the accessible cognitions inthe system (F(S, 104) = 4,13, /?"' = ,24, p < ,001). Results revealed that out of all thethoughts accessible to the ambivalent, only "There is a mountain of evidence" (fi =-, 15, p < .05) and "The court system long biased against Blacks "(fi = .24,p < ,02)predicted ambivalence.

To determine which CAUs played an independent role in bringing about sad-ness on the one hand, and happiness, on the other hand, in response to the verdict, weregressed ambivalence together with all the other accessible constructs in the pro-cessing system both on sadness and on happiness (F (9, 103) = 4.13, R^ = .24,p < ,001 and F(9, 103) = 3,63, R^=.24,p < ,001, respecdvely). Results indicatedthat the empathy the ambivalent felt for the vicdms and their families predicted thesadness they felt (/3 = .27, p < ,001), On die other hand, the recognidon that the ver-dict was a victory for Black people predicted their happiness (fi = ,26, p < .004),

Race artd Opinions of the Verdict

Correiadons between race and opinion of the verdict confirmed the popularview that being Black predicted agreement (r = ,54, p < ,001), whereas being Whitepredicted disagreement (r= ,5,p < ,001), In our sample, being Black also predictedambivalence toward the verdict (r= ,28,/>< ,03), To invesdgate whether race (spe-cifically, being Black or White) had an effect over and above the accessible CAUs indetermining one's opinion on the verdict, the regressions reported above were alsoconducted including race as an independent variable. When controlling for allCAUs operadng within each system, race was found to predict neither agreement (fi= ,15, ns), disagreement (fi = -.05, ns), nor ambivalence (fi = ,24, ns).

574 Meodoza-Denton, Ayduk, Sboda, and Mischel

Discussion

The results presented above support the idea that different cognidve-affecdvemediating tmits were accessible to p>eople who showed different reacdons to theSimpson verdict. Embedded within a network of acdvadon pathways, the set ofaccessible units consdtuted a subnetwork that self-sustained these units' acdvadonlevel. We illustrate these interacdons using cognidve-affecdve domain maps. AsMiller, Shoda, and Hurley (1996) have shown in the area of health-protecdvebehavior, these map>s allow us to tmderstand the processing dynamics that mediatebetween situadonal features in the environment and people's responses to them.

As these illustradons indicate, the CAPS model includes, but goes beyond,simple overall differences in the reacdons and emodons associated with each group.The model prompts us to take a closer look at what went on "inside the heads" ofindividuals as they processed informadon about the trial—as they thought about it,grappled with it, and resolved contradicdons within it. The domain maps show howpeople reacted in accordance with the prominent thoughts and feelings in their per-sonal worlds. In what follows, we trace the thought patterns among the dismayed,the elated, and the ambivalent, guiding the reader through each domain map. State-ments in quotadons are actual quotes from respondents in the study, and numbers inparentheses indicate corresponding arrow numbers within the maps being dis-cussed. Note that in discussing the domain maps, we often talk about one thoughtacdvadng or inhibidng another. We base these inferences on common sense,because our correladonal analysis does not allow us to determine the direcdonalityof acdvation.

The Dismayed

Figure 2 illustrates a CAPS that generated dismay in response to the verdict.The domain map shows that Simpson's escape in the Bronco, the physical evidence,and Simpson's past history of spousal abuse all posidvely activated thoughts that"O. J, Simpson is gtiilty, guilty, guilty" (1-4), These cognidons inhibited recogni-don of the argument that the evidence may have been quesdonable (5-7), Widi thesecognidons strongly activated, informadon related to Mark Fuhrman easily led tothoughts that "the trial should not have been turned into a debate on racism" (8),which further strengthened the idea that die evidence was solid (9), but also inhib-ited the idea that the evidence may have been tainted (10), In light of all the evidencethe prosecudon presented, some respondents could only "blame the stupidity of thejury,.. they were probably just dumb" (11) and conclude that "the jusdce system isclearly in need of some restructuring" (12), The defense was encoded as havingsp)ent "a lot of dme pulling dirty tricks" (13), which was related to thoughts that"American jusdce is for sale to the highest bidder" (14),

CAPS Analysis (tf Reactions to the Simpson Case 575

DISMAY

Fig. 2. CAPS a.ssociated with disagreement with die Simpson criminal trial verdict. Solid lines indicatepositive (activating) relations among units; broken lines depict negative (deactivating) relations. Anode's darkness indicates its strength of accessibility. Dashed nodes indicate inhibited thoughts in thesystem.

With ideas about the preponderance of the evidence, die jury's inability tounderstand it, and the defense's devious tactics all accessible, the verdict itself wasgreeted widi disagreement (15-19), and a sense that diis was in no way an expres-sion of jusdce (20), Seeing people's reacdons reminded the dismayed that "the factthat he was Black and famous in America exonerated him from wrongdoing" (21)and about his domesdc abuse (22), This further strengthened the idea that the verdictwas wrong (23), and led to dismay over it (24),

The Elated

Figure 3 shows a prototypical CAPS that generated eladon in response to theverdict The domain map makes it clear that the personal worlds of the elated and thedismayed differed substandally. Among the first of these differences are thethoughts related to the tapes of Nicole Brown Simpson's emergency calls. As onerespondent noted, "I as well as everyone else did hear the 911 tapes. I feel that eventhough they had their marital problems, that would not be a concrete reason to killsomeone" (1), Some respondents pointed out diat Nicole Brown Simpson "had

576 Mendoza-Denton, Ayduk, Shoda, and Mischel

ELATION

FTg. 3. CAPS associated with agreement with the Simpson criminal trial verdict. Solid lines indicatepositive (activating) relations among units: broken lines depict negative (deactivating) relations. Anode's darkness indicates its strength of accessibility. Dashed nodes indicate itihibited thoughts in thesystem.

always been using drugs, borrowing money, and drinking and sleeping around" (2).Mark Fuhrman, "aracist cop" (3), immediately rendered the evidence quesdonable:"[Fuhrman] was going home, and all of a sudden he decided to turn around and lookfor evidence and mysteriously found a bloody glove" (4), The crime scene, ratherthan acdvadng thoughts about the "mountain of evidence," led to quesdoning Simp-son's involvement in the crime. "Why would O. J,," wrote one pardcipant, "murderhis wife, and, as any sensible person realize he would be a suspect, hide die bloodyclothes in his own house?" (5), The idea that the evidence was quesdonable inhib-ited many of the thoughts that were highly accessible to the dismayed (6-10),

Thus the elated, many of whom were skepdcal of the evidence, tended to feelthat the verdict was just (11) and that the jury "made a brave decision" (12-13). Inpart because of "racist assumpdons about the stupidity and barbarian temperamentof Black men," the verdict was also seen as "not just victory for one man, but a winfor all people who have been prejudged." All of these thoughts led to even strongerskepdcism toward the evidence (14-18). The accessibility of these thoughts mayhelp explain why the elated felt that Marcia Clark unfairly "tried her damnest to con-vict another Black man" (19-21),

CAPS Analysis of Reactions to the Simpson Case 577

The domain map also shows that two emodonal reacdons disdnguished thosewho agreed with the verdict: eladon and relief. Eladon was associated with thesense that the acquittal was a victory for Black people (22). On the other hand,seeing Simpson acquitted, despite the prejudging of Black men, brought aboutrelief (23-24), even if he "knows a hell of a lot more than he's saying" (25),

The Ambivalent

Figure 4 shows a prototypical CAPS that generated ambivalence in response tothe verdict. The figure allows us to appreciate the conflicted state of mind of theambivalent. On die one hand, they felt that Simpson was guilty "because hisbehavior just after the bodies of Nicole and Goldman were discovered [was]extremely suspicious" and "because too many things point[ed] in that direcdon"(1-3). On the other hand, they recognized that "there was much reasonable doubt,Mark Fuhrman was just one of the reasonable doubts" (4). While they felt that thecase shouldn't have been "more about race than the murder" (5-6), diey recognizedthat the acquittal of "a Black man killing a White woman in a town that was noto-rious for hatred from cops toward Negroes" was "cause for celebradon" (7-10). The

ntf vT Prucaxiat H>v(v«n (CAPSl

ELATIONSADNESS

Fig, 4. CAPS associated with ambivalence toward the Simpson criminal trial verdict. Solid lines indicatepositive (activating) relations among units; broken lines depict negative (deactivating) relations. Anode's darkness indicates its strength of accessibility. Dashed nodes indicate inhibited thoughts in thesystem.

578 Mendoza-Oentnn, Ayduk, Shoda, and Miscfael

conflicts and contradicdons that revolved in the ambivalent mind (11-14) madeeither agreeing or disagreeing a difficult task (15-17).

These respondents' ambivalence is apparent not only in their cognidons, butalso in their emodonal reacdons, TTiey felt happy that jusdce for Black p)eople wascarried out (18), but the suffering of the victims and families (19) made them feelthat the "the whole diing is a very sad event that no trial can really undo,,, it is sadfor all pardes involved" (20), The words of one pardcipant seem to capture theambivalent group best: "It's his case that on the whole represents something to me. Itmade me feel happy and sad, jusdce and injusdce, most of all it drew color lines, itmade me feel uncomfortable to sp>eak of the case."

Conclusions

In this ardcle, our principal aim was to take our understanding of the reacdonsto the Simpson verdict beyond the BlackAVhite dichotomy. Guided by a unifying,state-of-the-art framework based on findings from social cognidve theory andresearch, we showed that it was not race per se that determined whether a piersonagreed or disagreed with the verdict, but rather the networks of cognidons andaffects that informadon about the trial acdvated within individuals (see Fairchild &Cowan, this issue). The results, illustrated in selected cognidve-affecdve domainmaps, depict the acdvadon networks that determined how individuals felt about theverdict. The fact that many Black people agreed with the verdict was seen as reflect-ing a culturally shared network where discriminadon is a "hot," chronically acces-sible issue resuldng from minority status in America (Murray, Kaiser, & Taylor, thisissue). Just as importandy, however, we have shown that independent of race, beingsensidve to issues of discriminadon made agreement with the verdict more likely. Inthe impassioned words of one respondent:

1 believe very strongly he is innocent I have been the victim of police lies and violence,so I can see it from the victim's [Simpson's] side. The prosecution should be indicted forconspiracy—they behaved miserably. They knew Fuhrman was racist. They put blood onsocks. They lied.

This is the voice of a man who agreed with die verdict, and it is the voice of a Whiteman. He reminds us that the prism through which we perceive the world is the resultof a lifedme of exp)eriences—some more painful than others, some private, andsome shared.

In a muldcultural society, such as the one where the Simpson trial played itselfout, life experiences can vary from cultural group to cultural group. The CAPSmodel provides a novel methodology for looking through the prism of other experi-ences and cultures. With its emphasis on process, the model allows us to go beyondstereotypic conceptualizadons of cultural differences toward a deepwr appreciadonof the beliefs, values, and interreladonships among them that members of a groupmay share (Nisbett & Cohen, 1996; Sh weder, 1991). It encourages us to understand.

CAPS Analysis ot Reactions to the Simpson Case 579

rather than judge, how other p)eople see the world and to think about those situadonsin which different people may behave similarly, and similar people differendy.

References

Alston, W. P. (1975). Traits, consistency and conceptual altematives for personality theory. Joumalforthe Theory of Social Behaviour. 5. 17-48.

Anderson, J. R. (1983). The architecture of cognition. Cambridge, MA: Harvard University Press.Anderson, J. R. (1988). A spreading activation theory of memory. In A. Collins & E. E. Smith (Eds.),

Readings in cognitive science: A perspective from psychology and artificial intelligence (pp.138-154). San Mateo, CA: Morgan Kaufmann Publishers, Inc.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. EnglewoodCliffs, NJ: Prentice-Hall.

Bargh, J. A. (1994). The four horsemen of automaticity: Intention, awareness, efficiency, and control asseparate issues in social cognition. In R. S. Wyer and T. K. Snill (Eds.), Handbook of social cog-nition, (2nd ed.. Vol. 1). Hillsdale. NJ: Erlbaum.

Block, J, (1995). A contrarian view of the five-factor approach to personality description. PsychologicalBulletin, 117, 187-215.

Bower, G. H. (1978). Interference paradigms for meaningful propositional memory. American Journalof Psychology, 91,575-585.

Bruner, J. S. (1957). On perceptual readiness. Psychological Review. 64, 123-152.Cantor, N., & Kihlstrom, J. F. (1987). Personality and social intelligence. Hilisdale, NJ: Erlbaum.Carver, C. S. & Scheier, M. F. (1981). Self-con.«:iousness and reactance. Journal of Research in Person-

ality. 15. 16-29.Cervone, D. (1991). The two disciplines of persotiality psychology. Psychological Science. 2.371-377.Dodge, K. A. (1986). A social itiformation processing model of social competence in children. Cogni-

tive Perspectives on Children's Social Behavioral Dexelopment: The Minnesota Symposium onChild Psychology, 18. 77-125.

Geenz, C. (1973). The interpretation of cultures: Selected essays. New York: Basic Books.Gordon, E. (1982). Culture and Ethnicity. In M. D. Levine, W. B. Cavey, A. C. Crocker, and R. Gross

(Eds.). Textbook of behavioral and developmental pediatrics (pp. 187-192). Philadelphia, PA:W. B. Sanders.

Higgins. E. T. (1989). Knowledge accessibility and activation: Subjectivity and suffering from uncon-scious sources. In J. S. Uleman and J. A. Bargh (Eds.), Unintended thought (pp. 75-123). NewYork: Guilford.

Higgins, E. T. (1990). Personality, social psychology, and person-sittiation relations: Standards andknowledge activation as a cotnmon language. In A. Pervin (Ed.), Handbook of personality: The-ory and research (pp. 301-338). New York: Guilford.

Higgins, E. T, & Chaires, W. M. (1980). Accessibility of interrelational constructs: Implications forstimulus encoding and creativity. Journal of Experimental Social Psychology, 51. 5-15.

Higgins, E, T, & King, G. (1981). Accessibility of social consUTicts: Information processing conse-quences of individual and contextual variability. In N. Cantor &J. Kihlstrom (Eds.), Personality,cognition, and social interaction (pp. 69-121). Hillsdale, NJ: Erlbaum.

Hinton, G. E., McClelland, J. L., & Rumelhart, D. E. (1986). Distributed representations. In D. E. Ru-melhart & J. L. McClelland (Eds.), Parallel distributed processing: Explorations in the micro-structures of cognition (Vol. 1: Foundations) (pp. 77-109). Cambridge, MA: MIT Press/Bradford Books.

Kihlstrom, J. F. (1990). The psychological unconscious. In L.A. Pervin (Ed.) Handbook of personality:Theory and re.iean:h (pp. 445-464). New York: Guilford Press.

Kluckhohi, C. (1965). Culture and behavior New York: Free Press.Kunda, Z., & Thagaid, P. (1996). Forming impressions from stereotypes, traits, and behaviors: A

parallel-constraint-satisfaction theory. Psychological Review. 103.284-308.

580 Mendoza-Denton, Ayduk, Shoda, and Mischel

Lampert, M. D., & Ervin-Tripp, S. M. (1993). Structured coding for the study of language and social in-teraction. In J. A. Edwards & M. D. Lampert (Eds.), Talking data: Transcription and coding indiscourse research (pp. 169-206). Hillsdale, NJ: Erlbaum.

Margolick, D. (1995, October 4). Jury clears Simpson in double murder: Spellbound nation divides onverdicL The New York Times, pp. A1, A10.

Miller, S. M., Shoda, Y, & Hurley, K. (1996). Applying cognitive social theory to health protective be-havior: Breast self-examination in cancer screening. Psychological Bulletin, 119, 70-94.

Mischel, W. (1973). Toward a cognitive social learning reconceptualization of personality. Psychologi-cal Revie\^, 80 (4). 252-283.

Mischel, W. (1990). Personality dispositions revisited and revised: A view after three decades. In L. A.Pervin (Ed.), Handbook ofpersonality: Theory and research (pp. 111 -134). New York: GuilfordPress.

Mischel, W., & Shoda, Y. (1994). Personality psychology has two goals: Mu.st it be two fields? P.rycho-logical Inquiry. 5, 156-159.

Mischel, W., & Shoda, Y (1995), A cognitive-affective system theory ofpersonality: Reconceptualizingsittiations, dispositions, dynamics and invariance in personality structure. Psychological Re\iew,102. 246-268.

Mischel, W., Shoda, Y, & Rodriguez, M. L. (1989). Delay of gratification in children. Science. 244.933-938.

Nisbett, R., cS; Cohen, D. (1996). Culture of honor: Tlie psychology of violence in the South. Boulder,CO: Westview.

Obeyeskere, G. (1981). Medusa's hair: An essav on personal symbols and religious experience. Chi-cago: University of Chicago Press.

Pervin, L. A. (1989). Goal concepts: Themes, issues, and questions. In. L. A. Pervin (Ed.), Goal con-cepts in personality and social psychology (pp. 473^-479). Hillsdale, NJ: Erlbaum.

Pervin, L. A. (1994). A critical analysis of current trait theory. Psychological Inquiry. 5. 103-113.Read, S. J., Jones, D. K., & Miller, L. C. (1990). Traits as goal-based categories: The importance of goals

in the coherence of dispositional categories. Journal of Personality and Social Psychology, 58.104&-1061.

Read, S. J , & Miller, L. C. (1993). Rapist or "regular guy": Explanatory' coherence in the construction ofmental models of others. Personality and Social Psychology' Bulletin, 19,526-540.

Read, S. J., & Miller, L. C. (Eds.). (In press). Connectionist Models of Social Reasoning and Social Be-havior. Mahwah, NJ: Erlbaum.

Ross, L. & Nisbett, R. E. (1991). The person and the situation: Perspectives of social psychology. NewYork: McGraw-Hill.

Rumelhart, D. E., & McClelland, J. L. (1986). Parallel distributed processing: Explorations in the mi-crostructure of cognition (Vol. 1: Foundations). Cambridge, MA: MIT Press/Bradford Books.

Shoda, Y.. & Mischel, W. (1996). Toward a unified, intra-individual dynamic conception of personality.Journal of Research in Personality, 30.414-428.

Shweder, R. A. (1991). Thinking through cultures: Expeditions in cultural psychology. Cambridge, MA:Harvard University Press.

Spellman, B. A., Ullman, J. B., & Holyoak, K. J. (1993). A coherence model of cognitive consistency:Dynamics of attitude change during the Persian Gulf war. youraaio/5ocM(/.jsHfj,49,147-165.

Thagard, P. (1989). Explanatory coherence. Behavioral and Brain Sciences, 12, 435-467.Tylor, E. B. (1871). The origins of culture. Gloucester, MA: Peter Smith.

RODOLFO MENDOZA-DENTON is currendy a doctoral candidate at ColumbiaUniversity, His research interests include stereotyping, cultural idendty, andself-reguladon,

OZLEM N. AYDUK is currendy a doctoral candidate at Columbia University, Herresearch interests focus on gender idendty and gender differences in indmate rela-donships.

CAPS Analysis of Reactions to the Simpson Case 581

YUICHI SHODA received his Ph.D, from Columbia University in 1990. He is cur-rendy an Assistant Professor of Psychology at the University of Washington,

WALTER MISCHEL, Niven Professor of Humane Letters in Psychology atColumbia University, was the recipient of a Disdnguished Scientific ContributionAward from the American Psychological Associadon in 1982, His longdme col-laboradon with Dr. Yuichi Shoda has recently resulted in the development of theirCognidve-Affecdve System Theory of Personality.

Appendix A: O, J. Simpson Sights and Sounds

1) O. J. Simpson's slow-speed car chase (June 17, 1994).2) CNN computer-animated movie of the crime scene: Nicole Brown Simpson's

body,3) CNN computer-animated movie of the crime scene: Add Ron Goldman's body.4) Excerpt from Christopher Darden's opening statement (January 24, 1995),5) Excerpt from Johnnie Cochran's opening statement (January 25, 1995),6) Detecdve Mark Fuhrman is cross-examined by the defense.7) Excerpt from Laura Hart McKinney's tesdmony.8) Nicole Brown Simpson's 911 call (1989).9) Nicole Brown Simpson's 911 call (1993),10) Marcia Clark argues for Simpson's guilt,11) F, Lee Bailey raises quesdons about Mark Fuhrman's integrity (January 23,

1995).12) Prosecudon's closing argument (excerpt) (September 26, 1995).13) Defense's closing argument (excerpt) (September 27, 1995).14) Announcement of verdict (October 3, 1995).15) Goldman family reacdon to verdict.16) Simpson family reacdon to verdict,17) Crowd, outside courtroom, reacts to verdict.18) Simpson, a free man, returns to his estate and embraces Al Cowlings,

Copyright of Journal of Social Issues is the property of Wiley-Blackwell and its content may not be copied or

emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.