83589917

7/28/2019 83589917

1/18

Social Cognition, Vol. 30, No. 6, 2012, pp. 652668

652

2012 Guilford Publications, Inc.

Preparation o this article was supported through unding by the Social Sciences and HumanitiesResearch Council o Canada (SSHRC) and the Canada Research Chairs (CRC) Program.

Correspondence concerning this article should be sent to Bertram Gawronski, Department oPsychology, The University o Western Ontario, Social Science Centre, London, Ontario, N6A 5C2,Canada. E-mail: [email protected].

GAWRONSKI

COGNITIVE CONSISTENCYASACORE MOTIVE

baCk To The fUTUre of DissonanCe Theory:

CogniTive ConsisTenCy as a Core moTive

Btm GwoskiThe University of Western Ontario

I is to o cogiti dissoc, Fstig (1957) dscibd cogi-

ti cosistc s pscoogic d tt is s bsic s g dtist. O t pst dcds, ow, t id o cogiti cosistcs co moti s b pcd b icsig ow ocs odissoc-td cgs i ttitds d tti ccots tt t-tibt sc cgs to mcisms o go-ds. T ct ticims t iig t id o cogiti cosistc s co moti, gigtt icosistc ss s pistmic c o os i os sstmo bis. Bcs icosistc c ot b sod i mtip ws,motitd soig c bis pocsss o icosistc sotio towddsid cocsios, tog motitd distotios costid b td o cogiti cosistc. T biqit o cosistc pocsss is i-sttd tog its o i ios istcs o tt-compstio (.g.,

ictim dogtio, s-ifctio, sstm jstifctio) d t isigtstt c b gid om cocptizig ios soci pscoogicpom i tms o cogiti cosistc (.g., pjdic-td bisstms, dispositio ic, stbiit o fst impssios).

Festingers (1957) theory o cognitive dissonance is arguably one o the most in-uential theories in the history o social psychology. The theory postulates thatinconsistent cognitions elicit an aversive state o arousal (i.e., dissonance), whichin turn produces a desire to reduce the underlying inconsistency and to maintaina state o consonance.1 Although Festinger was convinced that the psychologicalneed or cognitive consistency is as basic as hunger and thirst, several revisions

1. Although Festinger (1957) preerred the term dissonance over inconsistency, it seems useul todistinguish between the inconsistency o cognitive elements and the aversive eeling o dissonancethat is assumed to arise rom inconsistent cognitive elements (see Gawronski, Peters, & Strack, 2008).

7/28/2019 83589917

2/18

CogniTive ConsisTenCy as a Core moTive 653

o his original theory have raised doubts about the status o cognitive consistencyas a core motive. These revisions have been inspired by evidence suggesting thatthe phenomenon o dissonance-related attitude change may not be as general as it

has been assumed in the early days o dissonance theory (or reviews, see Brehm,2007; Harmon-Jones & Harmon-Jones, 2007). To account or these deviations, at-titude change in dissonance paradigms (e.g., Brehm, 1956; Festinger & Carlsmith,1959) has been reinterpreted in terms o various alternative processes that have astronger resemblance to mechanisms o ego-deense than to Festingers originaltheory (or a discussion, see Greenwald & Ronis, 1978).

The main goal o the current article is to provide a conceptual analysis o the ideathat cognitive consistency represents a core motive that is indeed as undamentalas claimed by Festinger (1957). The central argument is that inconsistency servesas an epistemic cue or errors in ones system o belies, thereby imposing a ubiq-uitous constraint on thinking and reasoning that goes ar beyond the well-knowndemonstrations o dissonance-related attitude change. In this spirit, the currentarticle also aims at illustrating how the basic notion o cognitive consistency canintegrate a wide range o social psychological phenomena that have rarely beenanalyzed rom a consistency perspective.

The phenomenology of CogniTive ConsisTenCy

WhaT IS COGnITIve COnSISTenCy?

I cognitive consistency is interpreted in a relatively broad manner, it is possible tosubsume a large set o constructs and phenomena under this label (or reviews, see

Gawronski & Strack, 2012; Proulx, Inzlicht, & Harmon-Jones, 2012). Thus, to avoidconceptual conusion with related constructs, it is important to clearly speciy howcognitive consistency should be understood in the context o the current analysis.Festingers (1957) original defnition states that two cognitive elements are incon-sistent, i one element ollows rom the opposite o the other. More ormally, thisdefnition can be restated as: x andy are dissonant i not-x ollows romy (p. 13),with x and y subsuming any knowledge, opinion, or belie about the environ-ment, about onesel, or about ones behavior (p. 3).

Although this defnition may seem relatively simple and straightorward, thereare two important aspects that need to be considered or a conceptual analysis ocognitive consistency. First, the basic defnition entails that inconsistency is de-

fned by logical relations between cognitive elements. Second, the specifcationo the involved cognitive elements implies that these elements have to be under-stood as propositions about states o aairs that are regarded as true or alse bythe individual (Gawronski & Strack, 2004). Together, these criteria distinguishcognitive consistency rom purely semantic relations between cognitive concepts.For example, although the concepts extraverted and introverted are semantic ant-onyms, the concepts themselves are not consistent or inconsistent in the sense oFestingers (1957) defnition, unless they are part o propositions that relate themto an object o reerence. In act, simultaneous activation o semantically oppo-site concepts would produce cognitive inconsistency only i they are part o twopropositions that reer to the same object (e.g., Sally is extraverted and Sally is intro-verted). However, simultaneous activation o the two concepts would not produce

7/28/2019 83589917

3/18

654 gaWronski

cognitive inconsistency i they reer to dierent objects (e.g., Sally is extraverted andHelen is introverted). Moreover, even i two semantically opposite concepts reerto the same object, they would still be consistent i one o the two propositions is

regarded as alse (e.g., afrming the validity oSally is extraverted and negating thevalidity oSally is introverted). Hence, the involved cognitive elements have to beunderstood as propositional beliefs about states o aairs, which are characterized

by (a) a propositional relation between concepts (e.g.,John is smart; smoking causescancer;Mary likes cats, etc.) and (b) the assignment o a positive or negative truthvalue to the described relation (i.e., the subjective belie that the proposition istrue or alse). Propositional belies can be either general i they reer to categorieso objects (e.g., Germans are organized) or specifc i they reer to individual objects(e.g., Bertram is disorganized). Thus, counter to Festingers (1957) concern with therelation between two cognitive elements, inconsistency is most oten the result omore than two propositional belies (e.g., Germans are organized; Bertram is disorga-nized; Bertram is German).

COnSISTenCy, aMBIvalenCe, anD FluenCy

The specifcation o cognitive consistency as reerring to the logical relation be-tween propositional belies not only excludes semantic relations between indi-vidual cognitive concepts; it also distinguishes cognitive consistency rom relatedconstructs, such as ambivalence and uency. Ambivalence occurs when the mentalrepresentation o an object includes both positive and negative attributes (Van Har-reveld, Van der Pligt, & De Liver, 2009). In this sense, ambivalence can be the resulto two (or more) propositional belies implying that a target object possesses a pos-

itive and a negative attribute (Petty, Briol, & Johnson, 2012). However, ambiva-lence resulting rom evaluatively incongruent belies may not necessarily producecognitive inconsistency, i the relevant attributes involve dierent dimensions. Forexample, a person may be perceived as positive on one attribute-dimension andnegative on another attribute-dimension (e.g., Fred is competent and Fred is cold),which may cause ambivalence toward this person. Such perceptions do not pro-duce cognitive inconsistency in the sense o Festingers (1957) defnition, becausethe ascription o one attribute does not imply the opposite o another one (e.g.,

Judd, James-Hawkins, Yzerbyt, & Kashima, 2005). Nevertheless, ambivalence andinconsistency may sometimes co-occur, i the ascribed attributes are opposites othe same dimension (e.g., Fred is warm and Fred is cold). Whereas in the ormer case,

there is no need to resolve any inconsistency because both attributes can be pres-ent at the same time, the inconsistency in the latter case would have to be resolved,or example by speciying the contexts in which each ascription is accurate (e.g.,Fred is warm at home and Fred is cold at work; see Rydell & Gawronski, 2009).

Another construct that is closely related to cognitive consistency is the concepto uency, which describes the cognitive eelings that arise during the processingo inormation. Although the two constructs may seem rather similar, they areconceptually distinct in that cognitive (in)consistency reers to the content o theprocessed inormation (what?), whereas (dis)uency reers to the ease o process-ing that inormation (how?). Nevertheless, the two may sometimes inuence eachother, or example when the uency o processing a particular proposition is usedto evaluate the veracity o that proposition (see Winkielman, Huber, Kavanagh,

7/28/2019 83589917

4/18


& Schwarz, 2012). Whereas uent processing o a proposition tends to increasesubjective perceptions o its validity, disuency has been shown to reduce judg-ments o validity (e.g., Reber & Schwarz, 1999). Thus, higher levels o uency can

produce cognitive inconsistency when uent processing increases the perceivedvalidity o two inconsistent propositions.

Why IS COGnITIve COnSISTenCy IMPOrTanT?

The above specifcation o cognitive consistency may seem rather cold and logi-cal, and thus very dierent rom the notion o a basic motivational orce. So, whatmakes cognitive consistency so important that it deserves to be described as a coremotive? The psychological signifcance o cognitive inconsistency becomes cleari one considers its unction in signaling potential errors in ones system o belies

(Quine & Ullian, 1978). Although consistency is insufcient to establish accuracy,inconsistency is an unambiguous cue or errors that require appropriate revisions.For example, i a person believes that (a) Good friends always support each other whenthey need help, (b) Lucy and Linda are good friends, and (c) Lucy did not support Lindawhen Linda needed help, the implied inconsistency requires a reassessment o thethree belies, which may lead to the revision o either one o them. It may lead toa revision o the belie that good riends always support each other by allowingor exceptions (e.g., Good friends may sometimes fail to support each other when theyhave their own problems); it may lead to a revision o the belie that Lucy is actuallya good riend o Linda (e.g., Lucy always claimed that she is a good friend of Linda,but thats not true); or it my lead to a revision o the belie that Lucy did not sup-port Linda when Linda needed help (e.g., Lucy offered to help Linda, but Linda did

not want Lucys help). From a pragmatic perspective, the identifcation o errorsin ones system o belies is important, because erroneous belies can underminecontext-appropriate behavior by suggesting inadequate courses o action (Quine& Ullian, 1978). Moreover, cognitive inconsistency itsel can sometimes disruptcontext-appropriate behavior, because inconsistent belies may suggest mutuallyexclusive courses o action (Harmon-Jones, Amodio, & Harmon-Jones, 2009). In

both cases, the state o arousal that is assumed to be elicited by inconsistent beliesserves as a signal that the current system o belies has to be revised to acilitatecontext-appropriate action.

hOW DO We IDenTIFy InCOnSISTenCy?

Although the identifcation o inconsistency may seem relatively straightorwardin the above examples, it can be quite difcult i it involves larger sets o belies.

Johnson-Laird (2012) pointed out that a set on propositions can be inconsistenteven i any n-1 o them yields a consistent set. The resulting capacity problem orthe identifcation o inconsistency is illustrated by the act that an exhaustive con-sistency assessment o a set o 100 propositions requires the consideration o 2100possibilities. Even i each possibility can be examined in a millionth o a second, acomprehensive examination would still take longer than the universe has existed.

Johnson-Laird argues that individuals resolve this capacity problem by means o

mental models. Instead o engaging in an exhaustive check o each combinatori-

7/28/2019 83589917

5/18

656 gaWronski

cally possible combination, an efcient way to evaluate the consistency o a givenset o propositional belies is to seek a mental model in which all the propositionsare true (Johnson-Laird, Girotto, & Legrenzi, 2004). I it is possible to fnd such a

model, the propositions will be judged consistent. I not, they will be judged asinconsistent.In addition to the combinatorical problem identifed by Johnson-Laird (2012),

limits in working memory capacity impose urther constraints by limiting thenumber o belies that can be simultaneously considered or an assessment oconsistency. Hence, the construction o mental models is also not comprehensive,

but limited to belies that are momentarily accessible (McGregor, Newby-Clark,& Zanna, 1999). An illustrative example is the hypocrisy paradigm (e.g., Stone,Aronson, Crain, Winslow, & Fried, 1994), in which participants are frst asked toindicate their general belies about a specifc issue in a pro-attitudinal manner(e.g., endorsement o the importance o sae sex) and then made aware o pastailures to behave in line with their general belies (e.g., recall o past ailures to usecondoms). The common fnding is that the inconsistency between general beliesand past behavior inuences subsequent behavior in a manner consistent with theendorsed belies (e.g., purchase o condoms). In other words, participants try toresolve the inconsistency between their general belies and their thoughts aboutpast behavior by changing their behavior. However, or this behavioral change tooccur, it is necessary that (a) general belies and thoughts about past behavior aresimultaneously accessible and (b) both o them are endorsed as valid. I one othe two conditions is not met, there will be no inconsistency identifed in the frstplace, and thus no change in behavior.

The Universal signifiCanCe of CogniTive ConsisTenCy

DOeS COGnITIve InCOnSISTenCy MaTTerOnly IF IT IS SelF-relevanT?

The most common way to test the implications o Festingers (1957) dissonancetheory is to investigate the conditions o dissonance-related attitude change. InFestinger and Carlsmiths (1959) induced compliance paradigm, or example, par-ticipants who engaged in counterattitudinal behavior tend to show a more avor-able evaluation o the relevant attitude object when they have received a low incen-tive or engaging in the counterattitudinal behavior than when they have received

a high incentive. The typical interpretation o this fnding is that a high incentiveresolves the aversive eeling o dissonance arising rom the inconsistency betweenthe attitude and the counterattitudinal behavior. In contrast, a low incentive isinsufcient to reduce the underlying inconsistency, thereby leading participants tochange their attitudes to bring them in line with their behavior. Another exampleis Brehms (1956) ree choice paradigm, in which participants tend to evaluate achosen object more avorably than a rejected object even i the two alternativeshave been evaluated equally beore the decision. The common explanation or thisspreading-of-alternativeseffect is that people experience an aversive eeling o post-decisional dissonance when they recognize that either (a) the rejected alternativehas positive eatures that the chosen alternative does not have, or (b) the chosen

alternative has negative eatures that are not present in the rejected alternative. In

7/28/2019 83589917

6/18


order to reduce this aversive eeling, people are assumed to emphasize or delib-erately search or positive characteristics o the chosen alternative and negativecharacteristics o the rejected alternative, which inuences their subsequent evalu-

ations o the two alternatives.Challenging the initial assumption that these phenomena reect the operation oa ubiquitous need or cognitive consistency, several subsequent studies have ailedto show dissonance-related attitude change under particular conditions. These

boundary conditions have led some researchers to conclude that the need or con-sistency is not as universal as proposed by Festinger (1957). Instead, dissonance-related attitude change was attributed to mechanisms o ego-deense, such as theneed to maintain consistent views about the sel (e.g., Aronson, 1968), the need tomaintain a positive sel-image (e.g., Steele & Liu, 1983), eelings o personal re-sponsibility or producing aversive consequences (e.g., Cooper & Fazio, 1984), orthe need to maintain sel-views that are consistent with ones personal standards(e.g., Stone & Cooper, 2001). Although the proposed reormulations dier in im-portant regards, they share the assumption that inconsistency matters only i it isrelated to the sel.

In line with pragmatic interpretations (Harmon-Jones et al., 2009; Quine & Ul-lian, 1978), I argue that cognitive consistency is much more undamental and ubiq-uitous than suggested by reormulations o dissonance theory that attribute con-sistency-related phenomena to mechanisms o ego-deense (see also Greenwald &Ronis, 1978; Proulx et al., 2012). An important aspect in this regard is that manyaccounts equate the presence versus absence o dissonance-related attitude changewith the presence versus absence o aversive eelings arising rom inconsistency.This equation is problematic, because it conates the behavioral phenomenonthat needs to be explained (i.e., attitude change) with the psychological mecha-

nism that is supposed to explain the observed phenomenon (i.e., inconsistencyresolution; or a more detailed discussion, see De Houwer, Gawronski, & Barnes-Holmes, 2012). Because cognitive inconsistency can be resolved in many waysother than attitude change (Festinger, 1957), the mere presence versus absence oattitude change does not allow any conclusions about cognitive inconsistency it-sel or the aversive eelings arising rom cognitive inconsistency. Hence, althoughthe observed boundary conditions provide important inormation about the mod-erators o attitude change as one o several strategies to resolve inconsistency, theconditions o a particular strategy to resolve inconsistency should not be conusedwith the presence versus absence o inconsistency itsel or the aversive eelingsresulting rom cognitive inconsistency.

COGnITIve COnSISTenCy aCrOSS CulTure, aGe, anD SPeCIeS

The past decades o dissonance research have been characterized by recurring de-bates about the presumed universality o cognitive consistency as a core motive.On the one hand, many cross-cultural researchers are convinced that dissonanceis a culture-specifc phenomenon that is limited to Western, individualist coun-tries and less likely in Eastern, collectivists countries (e.g., Heine & Lehman, 1997;Markus & Kitayama, 1991). On the other hand, researchers in developmental andcomparative psychology have argued that dissonance can also be observed in in-ants and various animals (e.g., Egan, Santos, & Bloom, 2007; Lydall, Gilmour, &

7/28/2019 83589917

7/18

658 gaWronski

Dwyer, 2010), which is consistent with Festingers (1957) original interpretation ocognitive consistency as a universal motive.

Although the current proposal agrees with the latter view, both perspectives

require a careul analysis o the available evidence to avoid premature conclu-sions. An important aspect in this regard is again the distinction between behav-ioral eects (e.g., attitude change) and the psychological mechanisms that pre-sumably underlie these eects (e.g., inconsistency resolution). In the cross-culturalliterature, or example, the absence o attitude change as a result o induced com-pliance (Festinger & Carlsmith, 1959) or ree choice (Brehm, 1956) in East Asianparticipants is oten interpreted as evidence that East Asians do not have a basicneed or cognitive consistency, and thereore do not experience aversive eelingso dissonance (e.g., Heine & Lehman, 1997; Markus & Kitayama, 1991). As notedabove, however, the absence o dissonance-related attitude change can be due tomultiple actors that do not imply the absence o a basic need or cognitive con-sistency. One important actor in this regard is the contextualized notion o truthin Eastern cultures, which diers rom the decontextualized way o thinking inWestern cultures. According to the latter, the truth or alsity o a given proposi-tion about an object is determined by the intrinsic properties o that object. Thisworld view stands in contrast to the dialectical way o thinking in many Easterncultures, according to which the truth or alsity o a proposition depends on theparticular context; it may be true in some regards, but the opposite may be truein other regards (Peng & Nisbett, 1999). Because a decontextualized acceptance orrejection o a given proposition involves a higher likelihood o inconsistency withother propositions, inconsistency is more likely within belie systems that involvedecontextualized belies. Yet, inconsistency is less likely within belie systems thatinvolve contextualized belies. Importantly, the latter does not imply that consis-

tency would be irrelevant; it simply means that what may be inconsistent roma Western, decontextualized point o view may be perectly consistent rom anEastern, contextualized point o view.

Another important aspect in the interpretation o cross-cultural dierences con-cerns the preerred strategy to resolve inconsistency. Several studies on causalattribution have shown that participants rom Eastern cultures have a strongertendency to explain behavior in situational terms than participants rom Westerncultures (e.g., Miller, 1984; Morris & Peng, 1994), and this tendency may reduce thelikelihood o dissonance-related attitude change. For example, whereas Easternersmay explain their counterattitudinal behavior in induced compliance situations(see Festinger & Carlsmith, 1959) by means o situational demands regardless o

whether situational pressure is high or low (e.g., The experimenter asked me to do it),Westerners may explain their counterattitudinal behavior in situational terms onlywhen situational pressure is high (e.g., The experimenter gave me $20 for doing it), butnot when situational pressure is low (e.g., The experimenter gave me $1 for doing it).Importantly, such dierences in attributional tendencies do not imply that incon-sistency is irrelevant in East Asian cultures; it simply means that inconsistency ismore likely to be resolved by means o situational attributions.

Although the current proposal agrees with developmental and comparativepsychologists who argue that the need or consistency is a universal phenomenonthat can even be observed in inants and animals, the evidence that is requentlycited in support o this claim also has to be interpreted with caution. Many othese studies relied on the ree choice paradigm (Brehm, 1956) to demonstrate the

7/28/2019 83589917

8/18


emergence o a spreading-o-alternatives eect in inants and animals (e.g., Eganet al., 2007). However, the mere demonstration o a spreading-o-alternatives e-ect is insufcient to establish the presence o cognitive inconsistency, because this

eect can be due to various other actors, including artiacts o the experimentaldesign (Chen & Risen, 2010). These conounds resonate with the above concernthat behavioral eects (e.g., attitude change) should not be conused with theirunderlying mental mechanism (e.g., inconsistency resolution).

Despite these concerns, it is important to note that responses to cognitive incon-sistency are a central component o developmental research with preverbal inants.In this research, enhanced visual attention is oten interpreted as an indicator oexpectancy violation to study various aspects o cognitive development. Such ex-pectancy violations can be interpreted as cases o cognitive inconsistency, in whicha prior belie about a particular state o aairs conicts with a new observation(or a similar approach in animal research, see Tinkelpaugh, 1928). Thus, althoughresearch demonstrating dissonance-like phenomena in animals and inants should

be interpreted with caution or the reasons outlined above, the need or cognitiveconsistency represents the oundation or a wide range o paradigms to study thecognitive development o preverbal inants. To be sure, animals and inants mayhave lower working memory capacity than human adults, which may constrainthe complexity o mental models to identiy inconsistency (see Johnson-Laird etal., 2004). However, this constraint does not imply the absence o a universal needor cognitive consistency. Ater all, any organism requires an accurate representa-tion o the world or context-appropriate action, and inconsistency serves as animportant signal o inaccurate representations.

COGnITIve COnSISTenCy anD MOTIvaTeD reaSOnInG

The current analysis is based on the premise that cognitive inconsistency servesas an important signal or errors in ones system o belies. However, a large bodyo research suggests that people are oten more concerned with what makes themeel good than what is accurate (or an overview, see Alicke & Sedikides, 2010). Forexample, Kruglanski and Shteynberg (2012) have argued that consistency is not amotivational orce in itsel, but the accidental outcome o epistemic processes thataim at validating propositions that are desired and invalidating propositions thatare undesired. The aective consequences o (in)consistency are urther assumedto depend on the nature o peoples goals, not on (in)consistency per se. Specif-

cally, consistency should elicit positive eelings when it validates a desired belie,but negative eelings when it validates an undesired belie. Conversely, inconsis-tency is assumed to elicit negative eelings when it invalidates a desired belie, butpositive eelings when it invalidates an undesired belie.

Although Kruglanski and Shteynbergs (2012) analysis highlights the impor-tance o considering the relation between cognitive consistency and motivatedreasoning, it conates the consistency o unqualifed propositions with the con-sistency o subjective propositional belies. As outlined at the beginning o thisarticle, cognitive consistency reers to propositions about states o aairs that areregarded as true or alse by the individual, not to the unqualifed content o propo-sitions independent o their subjective truth or alsity. O course, a person may bemotivated to invalidate an undesired proposition by identiying its inconsistency

7/28/2019 83589917

9/18

660 gaWronski

with another proposition that is regarded as true. In such cases, however, the con-tents o the relevant propositions are qualifed, in that the undesired proposition isrejected as alse whereas the invalidating proposition is regarded as true. Thus, it

is not the (in)consistency between unqualifed propositions that unctions as a coremotivational orce (c. Kruglanski & Shteynberg, 2012), but the (in)consistency oa persons subjective propositional belies. Using the above example o Lucy andLinda, a person might be motivated to believe in the unconditional support pro-vided by good riends, and thus search or inormation that questions the qualityo Lucys riendship. In this case, the sought-ater inormation would be incon-sistent with the unqualifed content o the proposition about Lucys riendshipto Linda. However, because the latter proposition is rejected as alse, the overallsystem o belies remains consistent, although motivational inuences clearly in-uenced the particular manner in which inconsistency has been resolved.

Another important aspect in this context is that inconsistency can oten be re-solved in multiple ways, leaving enough exibility or inuences o motivated rea-soning (Kunda, 1990). For example, whereas some peers o Lucy and Linda may

be motivated to maintain a positive impression o Lucy, others may be less con-cerned about that. Thus, the ormer group might selectively search or inorma-tion suggesting that Linda rejected Lucys oers or help, whereas the latter groupmight consider inormation that questions the quality o Lucys riendship. Yet,what is important in both cases is that the resulting systems o subjective beliesare consistent. Although many cases leave enough exibility to resolve cognitiveinconsistency in ways that involve the retention o desired belies and the rejectiono undesired belies, consistency imposes a universal constraint on any belie sys-tem. Ater all, motivated reasoning involves a motivation to believe in the truth odesired belies and the alsity o undesired belies, and inconsistency would signal

that there is an error in the overall system o belies.

COGnITIve COnSISTenCy anD ThreaT-COMPenSaTIOn

The proposed relation between cognitive consistency and motivated reasoningalso provides a conceptual integration o various instances o threat-compensa-tion. Specifcally, the current analysis suggests that many compensatory responsesto threatening inormation can be understood as cases in which newly acquiredinormation conicts with an existing belie (see also Proulx et al., 2012). To theextent that people are motivated to retain their prior belies, they may resolve the

implied inconsistency in a manner that confrms the subjective validity o thesebelies.

A useul example to illustrate the role o cognitive consistency in threat-com-pensation is the derogation o innocent victims as a result o just-world belies.The belie in a just world reects the conviction that the world is a just place inwhich people get what they deserve: good things happen to good people and badthings happen to bad people (Lerner, 1980). A common fnding in the literature on

just-world belies is that innocent victims tend be derogated or blamed or theirmisortune (e.g., Lerner & Simmons, 1966), and this tendency has been shown toincrease as a unction o just-world belies and to decrease when justice is restored.For example, women who were raped by a male perpetrator are oten accusedo being partially responsible or the crime (e.g., as implied by claims that they

7/28/2019 83589917

10/18


sexually provoked the perpetrator by the way they dressed), and such attributionso blame tend to increase as a result o just-world belies and decrease when theperpetrator is punished or his crime.

From a cognitive consistency perspective, one could argue that inormationabout an innocent victim creates a conict between the general belie that theworld is a just place in which people get what they deserve and the belie thatthe victim is a good person who did not deserve his or her misortune (Van denBos & Maas, 2009). To the extent that people are motivated to retain their belie ina just world (because it provides a sense o justice and airness), a potential wayto resolve this inconsistency is to reject the proposition that the victim is a goodperson who did not deserve his or her misortune. Moreover, i justice is restored(e.g., when the perpetrator o a crime is punished), the initial inconsistency is re-solved, thereby reducing the need to restore consistency by blaming or derogatingthe victim. Yet, to the extent that people do not belie in a just world to begin with,there is no inconsistency, and thereore no need to restore consistency by blamingor derogating the victim.

Similar considerations can be applied to other instances o threat-compensation.For example, deensive responses to negative inormation about the sel usuallyinvolve a conict between an existing positive belie about the sel (e.g., I am a goodstudent) and a newly acquired specifc belie (e.g., I got a low mark on the essay). Tothe extent that people are motivated to retain their prior views about themselves,they may resolve this inconsistency by searching or inormation that questionsthe diagnosticity o the newly acquired inormation (e.g., The instructor doesnt likeme). As demonstrated in research on sel-verifcation, however, such deensiveresponses are limited to cases in which people hold positive belies about them-selves. I their sel-belies are negative, there is no inconsistency to begin with,

thereby reducing the need to restore consistency by searching or discounting in-ormation. In act, people with negative sel-views usually accept the validity onegative eedback, but question the validity o positive eedback (or a review, seeSwann & Brooks, this issue). In either case, people seem to be motivated to retaintheir prior belies about the sel which leads them to restore inconsistency in amanner that questions the validity o newly acquired inormation that is inconsis-tent with their prior sel-view.

Another illustrative example is the tendency to justiy the current social ordereven i the status quo goes against ones personal interests (Jost, Banaji, & Nosek,2004). A common fnding in the literature on system justifcation is that the mem-

bers o disadvantaged groups oten adopt a negative stereotypical view o their in-

group, thereby protecting their belies about the airness o the current social struc-ture. From a cognitive consistency perspective, one could argue that such reactionsto social inequality have their roots in the conict between the general belie thatthe existing social structure is air and the specifc belie that ones ingroup is dis-advantaged. To the extent that people are motivated to retain their general belieabout the airness o the current system, they may restore consistency by adopt-ing the belie that the ingroup is inerior, and thereore deserves to be in a lowersocial position. Yet, i people reject the idea that the current system is air, there isno inconsistency to begin with, thereby eliminating the need to restore consistency

by derogating ones ingroup. Similarly, the status quo may be deended by com-plementary belies, such as ascriptions o positive and negative attributes alongdierent dimensions (e.g., poor but happy; see Kay, Jost, Mandisodza, Sherman,

7/28/2019 83589917

11/18

662 gaWronski

Petrocelli, & Johnson, 2007). In either case, processes o motivated reasoning biasthe resolution o inconsistency in a manner that helps to retain ones belies aboutthe airness o the existing social order, thereby protecting these belies against

potential counterevidence.

The UbiqUiTy of ConsisTenCy proCesses

Although cognitive consistency has been mostly studied in the context o disso-nance-related attitude change, its ubiquitous role is evident in a wide range ophenomena that are rarely considered as examples o consistency eects. The fnalsection o this article reviews three illustrative examples, showing that many socialpsychological phenomena can be understood in terms o consistency processes.These examples include the role o cognitive consistency in (a) prejudice-related

belie systems, (b) correspondent dispositional inerence, and (c) the stability ofrst impressions.

COGnITIve COnSISTenCy anD PreJuDICe

Over the past decades, research on prejudice has been guided by various con-structs, such as old-fashioned prejudice (Swim, Akin, Hall, & Hunter, 1995), modern

prejudice (McConahay, 1986), aversive prejudice (Dovidio & Gaertner, 2004), andimplicit prejudice (Rudman, Greenwald, Mellott, & Schwartz, 1999). However, theconceptual relation between these constructs has rarely been the subject o theoret-ical and empirical investigations. Recent work by Gawronski, Peters, Brochu, andStrack (2008) illustrates how a conceptualization in terms o cognitive consistencycan provide an integrative ramework or understanding the relation between theproposed constructs (or a review, see Gawronski, Brochu, Sritharan, & Strack,2012). A central assumption o their ramework is that automatic negative associa-tions elicit negative aective reactions toward targets o prejudice. These aectivereactions may serve as a basis or a negative propositional evaluation o the targetgroup, unless a negative evaluation is inconsistent with other prejudice-relevant

belies (see Gawronski & Bodenhausen, 2006, 2011), including perceptions o dis-crimination and egalitarianism-related, nonprejudicial goals. With regard to racialprejudice against Arican Americans, or example, the set o judgment-relevant

belies may include the ollowing three components:

(1) I dislike Arican Americans.(2) Arican Americans are a disadvantaged group.(3) Negative evaluations o disadvantaged groups are wrong.

These three propositions are inconsistent with each other in that they cannot beendorsed simultaneously without violating the basic notion o cognitive consis-tency. Thus, inconsistency may be resolved by rejecting either one o the threepropositions. First, a person may reject Proposition 1 by not relying on his or hernegative aective reaction as a basis or an evaluative judgment (e.g., I like Afri-can Americans). Because not relying on an aective reaction does not necessarily

7/28/2019 83589917

12/18


eliminate the aective state, such a rejection results in a conict between aec-tive states and endorsed belies, as it is implied in the concept o aversive preju-dice (Dovidio & Gaertner, 2004). Second, a person may reject Proposition 2 by

disavowing the proposition that Arican Americans are a target o discrimination(e.g.,African Americans are not a disadvantaged group). Such perceptions play a cen-tral role in the concept o modern prejudice, which involves the denial o contin-ued discrimination in modern societies (McConahay, 1986). Third, a person mayreject Proposition 3 by disagreeing with the general disapproval o discriminatory

behavior (e.g., Negative evaluations of disadvantaged groups are okay). Such a rejec-tion o egalitarianism-related, nonprejudicial goals resonates with the concept oold-ashioned prejudice, such that people may reject egalitarian values and openlysupport racial discrimination (Swim et al., 1995). Importantly, all o these cases are

based on the premise that automatic negative associations elicit a negative aec-tive reaction toward targets o prejudice, which can be equated with the concept oimplicit prejudice (Rudman et al., 1999). Integrating dierent concepts o prejudicein such a manner not only illustrates the signifcance o cognitive consistency inprejudice-related belie systems; it also implies novel predictions about the rela-tion between implicit prejudice and the explicit endorsement o prejudiced belies,which received empirical support in several independent studies (or a review, seeGawronski et al., 2012).

COGnITIve COnSISTenCy anD DISPOSITIOnal InFerenCe

The correspondence bias is defned as the tendency to draw correspondent dispo-sitional inerences rom situationally constrained behavior (or a review, see Gaw-

ronski, 2004). A requently cited example is the case o attitude attribution (Jones& Harris, 1967), in which perceivers tend to iner a corresponding attitude rom anendorsed position even when the author was not ree to chose the position (e.g.,when the author o an essay was assigned to deend a particular position). Thecommon interpretation o this fnding is that perceivers tend to underestimate therole o situational actors (e.g., the act that the author was not ree to chose theendorsed position) in drawing correspondent dispositional inerences (Gilbert &Malone, 1995).

Counter to this interpretation, Gawronski (2003) argued that attitude attribu-tions are guided by perceivers general belies about the ability to convincinglydeend an assigned position. To the extent that an endorsed position was reely

chosen by an author, this position is usually sufcient to iner that the author has acorresponding attitude. I, however, the endorsed position was assigned, perceiv-ers have to rely on alternative cues to iner the authors attitude. An importantcue in this regard is the persuasiveness o the authors arguments. Specifcally,perceivers seem to assume that only people with a corresponding attitude are ableto generate convincing arguments or an assigned position. In contrast, weak ar-guments may be generated by anyone regardless o their attitude (see Reeder &Brewer, 1979). Thus, when the endorsed position was not reely chosen by the au-thor, convincing arguments are highly diagnostic or inerring a corresponding at-titude. In contrast, unconvincing arguments remain ambiguous, in that the author

7/28/2019 83589917

13/18

664 gaWronski

may have an attitude that is either congruent or incongruent with the endorsedposition.

To test these assumptions, Gawronski (2003) told participants that the author

was asked to write (a) an essay in avor o a position that was determined random-ly or (b) an essay that was counter to his or her attitude. Consistent with the aboveassumptions, participants drew strong correspondent inerences rom randomlyassigned essays when the arguments in the essays were convincing, but not whenthey were unconvincing. More importantly, when participants were told that theauthor was assigned to write a counterattitudinal essay, unconvincing essays ledto strong inerences o attitudes that were opposite to the endorsed position, butthis eect was signifcantly reduced when the arguments in the essays were con-vincing. From a consistency perspective, one could argue that a strong essay con-icted with participants belie that only people with a corresponding attitude areable to write a persuasive essay. This inormation should lead them to expect arelatively weak essay when the author was ostensibly assigned to write a counter-attitudinal essay. Thus, i the essay turns out to be highly persuasive, perceiversare conronted with three inconsistent propositional belies (Gawronski, 2009):

(1) Only people with a corresponding attitude are able to write a persuasive essay.(2) The author has been assigned to write a counterattitudinal essay.(3) The essay is highly persuasive.

These three propositions are inconsistent, in that each o them is in contradictionwith the logical implication o the other two (or a similar analysis o moral attri-

bution, see Gawronski, 2004). Because Proposition 3 is based on relatively hardevidence (i.e., the available essay), this proposition may be quite robust against

rejection. This leaves the possibility that either Proposition 1 or Proposition 2 needto be rejected. That is, the author may have reused to write a counterattitudinalessay, or people are indeed able to write highly persuasive essays in avor o acounterattitudinal position. Because the available evidence is insufcient to de-termine which o the two possibilities is the correct one, the perceiver is let witha state o uncertainty resulting rom the inconsistency between the three proposi-tions. Hence, dispositional inerences and subjective confdence are signifcantlyreduced (Gawronski, 2003). This example demonstrates not only that the process-es underlying correspondent dispositional inerences are much more complex andsophisticated than it is implied by the widespread assumption that people under-estimate the role o situational actors (see Gawronski, 2004). It also illustrates the

pervasive role o consistency processes in a domain that has rarely been related tothe notion o cognitive consistency (or a notable exception, see Kruglanski & Klar,1987).

COGnITIve COnSISTenCy anD FIrST IMPreSSIOnS

A fnal example o how an application o consistency principles can provide deep-er insights into classic social psychological phenomena is the stability o frst im-pressions. Although the importance o frst impressions is widely accepted in lay

7/28/2019 83589917

14/18


psychology, it is not clear why frst impressions tend to be so persistent to revision.A recent study by Rydell and Gawronski (2009) suggests that new experiencesthat contradict ones frst impression o another individual become bound to the

context in which they were made. Specifcally, their fndings showed that counter-attitudinal experiences with another person inuence evaluative responses only inthe context in which these experiences were made, whereas initial attitudes con-tinue to guide evaluative responses in any other context. Drawing on basic prin-ciples o cognitive consistency, Gawronski, Rydell, Vervliet, and De Houwer (2010)argued that expectancy violations enhance perceivers attention to contextual cuesto identiy potential explanations or the observed deviation (see Roese & Sher-man, 2007). Thus, experiences that contradict a frst impression tend to be stored incontextualized representations, which leaves initially ormed context-ree repre-sentations intact. As a result, the content o expectancy-violating experiences will

become activated only in the particular context in which these experiences weremade. Yet, encountering the target in any other context, be it the context o the frstimpression or a novel context, should activate the initial context-ree representa-tion, thereby leading to a dominance o the frst impression. From a cognitive con-sistency perspective, one could argue that expectancy violations lead perceivers tosearch or a situational explanation or the unexpected experience to resolve theinconsistency between the expectancy and the expectancy-violating event. To theextent that the identifed cause is integrated into the representation o the newlyacquired inormation, the initial expectancy can be retained or any situation thatdiers rom the situation in which the expectancy has been violated. Interestingly,the contextualization o expectancy-violating experiences also prevents cognitiveinconsistency rom the simultaneous activation o conicting representations, be-cause the context modulates which representation will be activated in a given situ-

ation, the initial context-ree representation or the subsequently ormed contextu-alized representation.

ConClUsions

The main goal o the current article was to revive Festingers (1957) seminal hy-pothesis that cognitive consistency represents a core motive underlying thinkingand reasoning. I argued that cognitive inconsistency serves as an epistemic cueor errors in ones system o belies that require a reassessment o their validity.Because there are always multiple ways to resolve inconsistency, motivated rea-

soning can bias belie systems toward desired conclusions, although the need orconsistency imposes a basic constraint on any such distortions. Acknowledgingthe undamental nature o cognitive consistency not only opens the door or a con-ceptual integration o various instances o threat-compensation; it also providesnovel insights into a wide range o phenomena that have rarely been studied roma consistency perspective.

7/28/2019 83589917

15/18

666 gaWronski

referenCes

Alicke, M. D., & Sedikides, C. (Eds.). (2010).

Handbook of self-enhancement and self-pro-tection. New York: Guilord.

Aronson, E. (1968). Dissonance theory: Prog-ress and problems. In R. P. Abelson, E.Aronson, W. J. McGuire, T. M. New-combe, M. J. Rosenberg, & P. H. Tan-nenbaum (Eds.), Theories of cognitive con-sistency: A sourcebook (pp. 5-27). Skokie,IL: Rand-McNally.

Brehm, J. W. (1956). Postdecision changes inthe desirability o alternatives.Journal of

Abnormal and Social Psychology, 52, 384-389.

Brehm, J. W. (2007). A brie history o disso-nance theory. Social and Personality Psy-chology Compass, 1, 381-391.

Chen, M. K., & Risen, J. L. (2010). How choiceaects and reects preerences: Revisit-ing the ree-choice paradigm. Journal ofPersonality and Social Psychology, 99, 573-594.

Cooper, J., & Fazio, R. H. (1984). A new lookat dissonance theory.Advances in Experi-mental Social Psychology, 17, 229-266.

De Houwer, J., Gawronski, B., & Barnes-Holmes, D. (2012).A functional-cognitive

framework for attitude research. Manu-script submitted or publication.

Dovidio, J. F., & Gaertner, S. L. (2004). Aversiveracism. Advances in Experimental SocialPsychology, 36, 1-52.

Egan, L. C., Santos, L. R., & Bloom, P. (2007).The origins o cognitive dissonance:Evidence rom children and monkeys.Psychological Science, 18, 978-983.

Festinger, L. (1957). A theory of cognitive disso-nance. Evanston, IL: Row Peterson.

Festinger, L., & Carlsmith, J. M. (1959). Cogni-tive consequences o orced compliance.

Journal of Abnormal and Social Psychology,58, 203-210.

Gawronski, B. (2003). Implicational schemataand the correspondence bias: On thediagnostic value o situationally con-strained behavior. Journal of Personalityand Social Psychology, 84, 1154-1171.

Gawronski, B. (2004). Theory-based bias cor-rection in dispositional inerence: Theundamental attribution error is dead,long live the correspondence bias. Eu-

ropean Review of Social Psychology, 15,

183-217.Gawronski, B. (2009). The multiple inerence

model o social perception: Two con-ceptual problems and some thoughtson how to resolve them. PsychologicalInquiry, 20, 24-29.

Gawronski, B., & Bodenhausen, G. V. (2006).Associative and propositional processesin evaluation: An integrative review oimplicit and explicit attitude change.Psychological Bulletin, 132, 692-731.

Gawronski, B., & Bodenhausen, G. V. (2011).The associative-propositional evalua-

tion model: Theory, evidence, and openquestions. Advances in Experimental So-cial Psychology, 44, 59-127.

Gawronski, B., Brochu, P. M., Sritharan, R., &Strack, F. (2012). Cognitive consistencyin prejudice-related belie systems: In-tegrating old-ashioned, modern, aver-sive and implicit orms o prejudice. InB. Gawronski & F. Strack (Eds.), Cogniti-ve consistency: A fundamental principle insocial cognition (pp. 369-389). New York:Guilord.

Gawronski, B., Peters, K. R., Brochu, P. M., &

Strack, F. (2008). Understanding the re-lations between dierent orms o racialprejudice: A cognitive consistency per-spective. Personality and Social Psycholo-

gy Bulletin, 34, 648-665.Gawronski, B., Peters, K. R., & Strack, F. (2008).

Cross-cultural dierences vs. universal-ity in cognitive dissonance: A concep-tual reanalysis. In R. M. Sorrentino & S.Yamaguchi (Eds.), Handbook of motivati-on and cognition across cultures (pp. 297-314). New York: Elsevier.

Gawronski, B., Rydell, R. J., Vervliet, B., & De

Houwer, J. (2010). Generalization versuscontextualization in automatic evalua-tion. Journal of Experimental Psychology:General, 139, 683-701.

Gawronski, B., & Strack, F. (2004). On the prop-ositional nature o cognitive consisten-cy: Dissonance changes explicit, but notimplicit attitudes.Journal of ExperimentalSocial Psychology, 40, 535-542.

Gawronski, B., & Strack, F. (Eds.). (2012). Co-gnitive consistency: A fundamental princip-le in social cognition. New York: Guilord.

7/28/2019 83589917

16/18


Gilbert, D. T., & Malone, P. S. (1995). The cor-respondence bias. Psychological Bulletin,117, 21-38.

Greenwald, A. G., & Ronis, D. L. (1978). Twen-

ty years o cognitive dissonance: Casestudy o the evolution o a theory. Psy-chological Review, 85, 53-57.

Harmon-Jones, E., Amodio, D. M., & Harmon-Jones, C. (2009). Action-based model odissonance.Advances in Experimental So-cial Psychology, 41, 119-166.

Harmon-Jones, E., & Harmon-Jones, C. (2007).Cognitive dissonance theory ater 50years o development. Zeitschrift fr So-

zialpsychologie, 38, 7-16.Heine, S. J., & Lehman, D. R. (1997). Culture,

dissonance, and sel-afrmation. Perso-nality and Social Psychology Bulletin, 23,389-400.

Johnson-Laird, P. N. (2012). Mental modelsand consistency. In B. Gawronski & F.Strack (Eds.), Cognitive consistency: A

fundamental principle in social cognition(pp. 225-244). New York: Guilord.

Johnson-Laird, P. N., Girotto, V., & Legrenzi,P. (2004). Reasoning rom inconsistencyto consistency. Psychological Review, 111,640-661.

Jones, E. E., & Harris, V. A. (1967). The attribu-

tion o attitudes. Journal of ExperimentalSocial Psychology, 3, 1-24.Jost, J. T., Banaji, M. R., & Nosek, B. A. (2004).

A decade o system justifcation theory:Accumulated evidence o conscious andunconscious bolstering o the statusquo. Political Psychology, 25, 881-919.

Judd, C. M., James-Hawkins, L., Yzerbyt, V.,& Kashima, Y. (2005). Fundamental di-mensions o social judgment: Under-standing the relations between judg-ments o competence and warmth.Jour-nal of Personality and Social Psychology,

89, 899-913.Kay, A. C., Jost, J. T., Mandisodza, A. N., Sher-man , S. J., Petrocelli, J. V., & Johnson,A. L. (2007). Panglossian ideology inthe service o system justifcation: Howcomplementary stereotypes help us torationalize inequality. Advances in Ex-

perimental Social Psychology, 39, 305-358.Kruglanski, A. W., & Klar, Y. (1987). A view

rom a bridge: Synthesizing consistencyand attribution paradigms rom a layepistemic perspective. European Journalof Social Psychology, 17, 211-241.

Kruglanski, A. W., & Shteynberg, G. (2012).Cognitive consistency as means to an end:

How subjective logic affords knowledge. InB. Gawronski & F. Strack (Eds.), Cogniti-

ve consistency: A fundamental principle insocial cognition (pp. 245-264). New York:Guilord.

Kunda, Z. (1990). The case or motivated rea-soning. Psychological Bulletin, 108, 480-498.

Lerner, M. J. (1980). The belief in a just world:A fundamental delusion. New York: Ple-num.

Lerner, M. J., & Simmons, C. H. (1966). Observ-ers reaction to the innocent victim:Compassion or rejection?Journal of Per-sonality and Social Psychology, 4, 203-210.

Lydall, E. S., Gilmour, G., & Dwyer, D. M.(2010). Rats place greater value on re-wards produced by high eort: An ani-mal analogue o the eort justifcationeect.Journal of Experimental Social Psy-chology, 46, 1134-1137.

Markus, H., & Kitayama, S. (1991). Cultureand the sel: Implications or cognition,emotion, and motivation. PsychologicalReview, 98, 224-253.

McConahay, J. B. (1986). Modern racism, am-bivalence, and the modern racism scale.

In J. D. Dovidio & S. L. Gaertner (Eds.),Prejudice, discrimination, and racism (pp.91-125). Orlando, FL: Academic Press.

McGregor, I., Newby-Clark, I. R., & Zanna, M.P. (1999). Remembering dissonance:Simultaneous accessibility o inconsis-tent cognitive elements moderates epis-temic discomort. In E. Harmon-Jones& J. Mills (Eds.), Cognitive dissonance:Progress on a pivotal theory in social psy-chology (pp. 325-353). Washington, DC:American Psychological Association.

Miller, J. G. (1984). Culture and the develop-

ment o everyday social explanation.Journal of Personality and Social Psycholo-gy, 46, 961-978.

Morris, M. W., & Peng, K. (1994). Culture andcause: American and Chinese attribu-tions or social and physical events.

Journal of Personality and Social Psycholo-gy, 67, 949-971.

Peng, K., & Nisbett, R. E. (1999). Culture, dia-lectics, and reasoning about contradic-tion.American Psychologist, 54, 741-754.

Petty, R. E., Briol, P., & Johnson, I. (2012). Im-plicit ambivalence. In B. Gawronski &

7/28/2019 83589917

17/18

668 gaWronski

F. Strack (Eds.), Cognitive consistency: Afundamental principle in social cognition(pp. 178-201). New York: Guilord.

Proulx, T., Inzlicht, M., & Harmon-Jones, E.

(2012). Understanding all inconsistency-compensation as a palliative response toviolated expectations. Trends in Cogniti-ve Sciences, 16, 285-291.

Quine, W. V. O., & Ullian, J. S. (1978). The web ofbelief(2nd ed.). New York: McGraw-Hill.

Reber, R., & Schwarz, N. (1999). Eects o per-ceptual uency on judgments o truth.

Consciousness and Cognition, 8, 338-342.Reeder, G. D., & Brewer, M. B. (1979). A sche-

matic model o dispositional attributionin interpersonal perception. Psychologi-cal Review, 86, 61-79

Roese, N. J., & Sherman, J. W. (2007). Expec-tancies. In E. T. Higgins & A. W. Krug-lanski (Eds.), Social psychology: Handbookof basic principles (2nd ed., pp. 91-115).New York: Guilord.

Rudman, L. A., Greenwald, A. G., Mellott, D.S., & Schwartz, J. L. K. (1999). Measuringthe automatic components o prejudice:Flexibility and generality o the ImplicitAssociation Test (IAT). Social Cognition,17, 437-465.

Rydell, R. J., & Gawronski, B. (2009). I like

you, I like you not: Understanding theormation o context-dependent auto-matic attitudes. Cognition and Emotion,23, 1118-1152.

Steele, C. M., & Liu, T. J. (1983). Dissonanceprocesses as sel-afrmation. Journal ofPersonality and Social Psychology, 45, 5-19.

Stone, J., Aronson, E., Crain, A. L., Winslow, M.P., & Fried, C. B. (1994). Inducing hypoc-

risy as a means o encouraging youngadults to use condoms. Personality andSocial Psychology Bulletin, 20, 116-128.

Stone, J., & Cooper, J. (2001). A sel-standards

model o cognitive dissonance. Journalof Experimental Social Psychology, 37, 228-243.

Swann, J. B., & Brooks, M. (2012). Why threatstrigger compensatory reactions: Theneed or coherence and quest or sel-verifcation. Social Cognition, 30, 758-777.

Swim, J. K., Aikin, K. J., Hall, W. S., & Hunter,B. A. (1995). Sexism and racism: Old-ashioned and modern prejudices. Jour-nal of Personality and Social Psychology,68, 199-214.

Tinkelpaugh, O. L. (1928). An experimentalstudy o representative actors in mon-keys. Journal of Comparative Psychology,8, 197-236.

Van den Bos, K., & Maas, M. (2009). On thepsychology o the belie in a just world:Exploring experiential and rationalisticpaths to victim blaming. Personality andSocial Psychology Bulletin, 35, 1567-1578.

Van Harreveld, F., Van der Pligt, J., & De Liver,Y. (2009). The agony o ambivalenceand ways to resolve it: Introducing theMAID model. Personality and Social Psy-

chology Review, 13, 45-61.Winkielman, P., Huber, D. E., Kavanagh, L.,& Schwarz, N. (2012). Fluency o con-sistency: When thoughts ft nicely andow smoothly. In B. Gawronski & F.Strack (Eds.), Cognitive consistency: A

fundamental principle in social cognition(pp. 89-111). New York: Guilord.

7/28/2019 83589917

18/18

Copyright of Social Cognition is the property of Guilford Publications Inc. 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.

83589917

Documents