Projection as an Interpersonal Influence Tactic

Jurnal:

Rucker, D. D., & Pratkanis, A. R. (2001). Projection as an interpersonal influence tactic: The effects of the pot calling the kettle black. Personality and Social Psychology Bulletin, 27(11), 1494-1507.

Note: Ini hanya sebuah catatan pribadi, mohon rujuk sumber asli

Projection as an Interpersonal Influence Tactic: The Effects of the Pot Calling the Kettle Black

Derek D. Rucker

Ohio State University

Anthony R. Pratkanis

University of California, Santa Cruz

In his analysis of Nazi propaganda, Ellis Freeman identified an influence tactic based on projection: accusing another person of the negative traits and behaviors that one possesses and exhibits with the goal of deflecting blame away from one’s own misdeeds and toward the accused. Although the use of projection as an influence tactic is not limited to fascist regimes and its consequences can be socially devastating, the projection tactic has not been subjected to experimental analysis. In four experiments, the authors found that projection was effective in (a) increasing the blame placed on the target of the projection and (b) decreasing the culpability of the accuser (or projectionist). These effects occurred despite (a) raising suspicions about the motives of the projectionist, (b) providing evidence that the projectionist is guilty of the deed, and (c) timing the projection so that it occurred after the misdeeds came to light. In February 1513, Niccolo Machiavelli, then secretary of Florence, was falsely accused of plotting to assassinate members of the ruling Medici family (de Grazia, 1994). Machiavelli lost his position, lost his Florentine citizenship, and was subsequently imprisoned and tortured for a crime he never committed. Ironically, Machiavelli’s accusers were the leading aristocrats of the times, all vying to replace the Medici family as leader of Florence. The experience so embittered Machiavelli that he wrote The Prince, perhaps the most cynical political treatise ever produced.

During the 1930s and early 1940s, Adolf Hitler wrongly accused the Jews of a plot to erect a despotism of Jewish world finance and to brutally dominate Europe and then the world (Freeman, 1940). Of course, when Hitler became the Führer of Germany, this is exactly what he attempted to do by trying to subjugate the rest of Europe and usher in the Third Reich of Nazi world domination.

In the 1950s, Senator Joseph McCarthy of Wisconsin conducted an investigation of alleged communist activities in the United States (Rovere, 1959). He would often accuse his opponents of the very characteristics he would then exhibit. For example, he called the publisher Henry Luce a liar and then proceeded to construct lies about Luce. In fact, he would challenge his opponents to take a lie detector test as he, McCarthy, was telling one lie after another. He denounced General George C. Marshall as a traitor who would sell his grandmother for any advantage, as McCarthy himself gained publicity for defaming this American hero. McCarthy’s attacks had a chilling effect on freedom of speech in the United States and devastated those he attacked. For example, General Marshall, the man associated with the plan that restored much of Europe from Hitler’s ruins, resigned his position as Secretary of Defense and never again served his nation.

Authors’ Note: This article is based on a senior honor’s thesis conducted by the first author under the direction of the second author.

This research was awarded the 1998 State University of New York, Stony Brook, prize for the best undergraduate research in political psychology and also received honorable mention for the 2000 SPSP student publication award. Preparation of this article was aided, in part, by a National Institute of Mental Health training grant (5 T32 MH-19728-08). We thank Kyle Rucker and Brenda Jackson for serving as confederates; Marlene E. Turner for assistance in data collection; Faye J. Crosby, G. William Domhoff, Anthony G. Greenwald, Richard E. Petty, Brad J.

Sagarin, M. Brewster Smith, and Michael J. Wenger for helpful comments; and Stacy Rucker for serving as a confederate and for helpful comments. Correspondence concerning this article should be sent to Derek D. Rucker, Department of Psychology, Ohio State University, 1885 Neil Avenue Mall, Columbus, OH 43210-1222, or Anthony R. Pratkanis, Department of Psychology, University of California, Santa Cruz, CA 95064; e-mail: rucker.46@osu.edu or peitho@cats.ucsc.edu.

PSPB,Vol. 27 No. 11, November 2001 1494-1507

© 2001 by the Society for Personality and Social Psychology, Inc.

­

Each of these three examples from history illustrates what we call the use of projection as an influence tactic. The concept is based on Ellis Freeman’s (1940) analysis of fascist and totalitarian propaganda, which often involved projecting the shortcomings of the leader (or group) onto a scapegoat. This projection would then serve the purpose of energizing and mobilizing the masses and of justifying extreme actions such as war or persecution of an outgroup.

The projection tactic is defined as accusing another of the same negative traits, characteristics, and behaviors that one possesses and exhibits with the goal of deflecting blame away from one’s own misdeeds. Recall that in each of our examples, the accusers (Florentine aristocrats, Hitler, and McCarthy) projected their own misdeeds (treason, world domination, lying) onto others.

This resulted in negative consequences for the accused:

Machiavelli was imprisoned, many European Jews were sent to internment camps and gas chambers, and General George C. Marshall quit politics. In addition, each of the accusers benefited from making their accusations.

The Florentine aristocrats were never suspected of or tried for treason; Hitler was supported by the German people, and his accusations bought time for his own military campaign; and McCarthy gained publicity. In each of these cases, projection had two complementary effects: First, it increased the blame placed on the accused (the target of projection), and second, it exonerated the accuser (or projectionist).

The use of projection as an influence tactic is not limited to political figures. For example, a child who has stolen a cookie, rather than accepting the parent’s punishment, may accuse a sibling of doing the deed. A spouse, guilty of having an affair, accuses her or his spouse of being disloyal to cover improprieties. A student, having cheated on a test, accuses another student of cheating to avoid suspicion from the teacher. Or consider a roommate who steals a friend’s wallet and when confronted accuses a mutual friend of the theft. Given these examples, it seems that projection may be a commonplace phenomenon, but just how effective is it as an influence device?

The purpose of this research is twofold. First, we seek to demonstrate the two complementary effects of the projection tactic: Projection is an effective means for (a) placing the accused in a negative light and (b) avoiding a negative evaluation of one’s own misbehavior. Support for the first hypothesized effect of projection, lowering the reputation of the accused, can be found in a series of experiments by Wegner, Wenzalaff, Kerker, and Beattie (1981). This research demonstrates that a negative rumor about a political candidate (delivered via a headline in a newspaper) can seriously damage the reputation of the rumor’s target regardless of the credibility of the source or validity of the rumor.

However, support for the second hypothesized effect of projection, exoneration of the projectionist, has not, to our knowledge, been demonstrated in previous work.

The second goal of this research is to identify boundary conditions for the interpersonal projection effect. In other words, are there factors that will increase or decrease the effectiveness of the projection tactic? In this research, we investigated the following variables that might be expected to limit the effectiveness of projection as an influence device: (a) suspicions about the motives of the accuser, (b) lack of ambiguity concerning the alleged misdeed, and (c) timing of the projection or accusation so that it occurs after the misdeed comes to light.

In our research, we are primarily concerned with the interpersonal effects of projection: what happens to the reputations of those involved in the giving and receiving of accusations. At this point, we are less concerned with the intrapsychic motivation for projection (an issue of importance for the clinical psychologist) and assume that a false accusation can occur for a variety of reasons, including attempts to rationalize a wrongdoing, deny a negative character trait (see Newman, Duff, & Baumeister, 1997), deal with frustration and anxiety, or be used in a strategic and calculated attempt to gain an advantage (as in political campaigns or organizational settings). In other words, we view our experiments as a snapshot of one component (the false allegation) in the complex and multifaceted rumor transmission process (see Fine, 1992, for details).

EXPERIMENT 1: A DEMONSTRATION OF THE PROJECTION TACTIC IN A THREE-PERSON COMPETITIVE GAME

Our first experiment attempted merely to establish the existence of the projection effect. Specifically, we sought to demonstrate that accusing others of a misdeed results in a (a) negative evaluation of the accused and (b) positive evaluation of the projectionist. In our first experiment, participants watched three people playing a prisoner’s dilemma type game. In this game, it is clear that one player is competing (and winning), despite telling the other players that he or she plans to cooperate on every trial. Half of the participants see and hear one of the players (Player B) accuse the others of lying and cheating (the projection tactic). This accusation is given twice by Player B, once at the middle of the game and once at the end of the game. Thus, we were able to investigate whether repeated accusations increase the effects of projection (as confident repetition of accusations leads to more acceptance) or decrease these effects (as observers become wary of the projectionist who cried wolf once too often). In this, our first experiment, we kept suspicions of Player B’s motives to a minimum, thus increasing the likelihood of demonstrating the projection effect. Experiment 1 can thus be seen as an analog to accusations made in the mass media (and other situations) where observers have little information and ability to discover the true facts of the case.

Method

PARTICIPANTS AND DESIGN

Participants included 58 University of California undergraduates who were randomly assigned in a 2 (projection: accusation vs. no accusation) 3 (game player:

A, B, and C) 2 (time of measurement: after the first accusation and after the second accusation) design (with game player and time of measurement serving as within subject factors). The students received credit in their introductory psychology class for their participation. ¹

PROCEDURE

Participants were recruited from the university participant pool for an experiment on “behavioral cues.” After arriving at the laboratory, participants were told that they would watch other participants from a previous experiment play 12 trials of a game called “prisoner’s dilemma.” (In actuality, they watched a videotape of a staged game.) The participant’s task was to rate the performance of those playing the prisoner’s dilemma game.

After completing informed consent materials, participants received more information about the game that they were to watch. In this game, three “former participants” who were called Player A, B, or C (to ensure anonymity) could make one of two decisions: cooperate or compete. If all three players cooperated, then everyone received one point; if two or more competed, then no one received any points; however, if one player competed and two cooperated, then the competitor received two points while everyone else received no points. In addition, participants were told that each player made both a public and a private decision. The public vote did not affect scoring but was used to indicate their intent to the other players. The private vote, however, was used for scoring purposes. Participants were informed that they would receive the actual scores of the game after the 6^th and 12th trials but that the scores would be given from high to low and not linked to a specific player. This was done so that participants would be suspicious, because there would be a high score, but they did not have evidence as to just which player was lying. Participants were told and encouraged to keep notes during the trials.

(Most participants took limited notes. For the most part, these notes reported facial expressions and other nonverbal behavior of the players.) Participants were informed that after the 6th and 12th trials, they would be asked to rate the strategy and playing style of each player.

To increase the credibility of the cover story, participants were told that they would be watching one of eight different groups of people playing the prisoner’s dilemma game. They were told that to prevent experimenter bias, the experimenter did not know which of the eight tapes they would be watching and that they would need to record the tape number when it was shown on the screen. The experimenter instructed the participant how to use the VCR and noted what buttons to push to start and stop the tape to complete the questionnaires at the end of the 6th and 12th trials. The experimenter then left the room.

Each participant next watched a videotape that began with an initial screen identifying the tape number. In actuality, all participants watched the same videotape.

This screen then faded out, and the prisoner’s dilemma game began with three players sitting around a table and engaged in social chatting. The players sat in a half circle with two women on the ends and a man (the projectionist) in the middle. The game was moderated by the same experimenter (the first author) who had introduced the experiment to the participant; he could not be seen on the videotape, but his voice could be heard instructing the players.

On each trial, players indicated a public vote by verbally saying, “cooperate” or “compete” and displaying a voting card with a halo (for cooperate) or a pitchfork (for compete). On all trials, each player displayed a halo and indicated that he or she would cooperate. After each player had indicated a public vote, a computer mouse was passed around the table for players to make a private vote. The private votes were not shown. During the trials of the game, communication by players was kept to a minimum by having players look straight ahead and only utter “cooperate,” “I will cooperate,” or some variation.

At the end of the sixth trial, the experimenter on the videotape announced that the scores were to remain confidential. Then, the screen went black, and these scores appeared on the TV screen: 10, 2, and 2, along with the caption “from high to low.” (This indicated that one person was competing whereas the other two were cooperating.) Participants in the projection treatment saw and heard Player B (the male player) accuse the other players of lying just before these scores appeared on the screen. Participants were then instructed to stop the VCR tape and complete the first questionnaire.

After completing the questionnaire, participants started the tape and watched the remaining trials, which proceeded in the same general manner as the first set of trials. At the end of the 12th trial, the score report was once again given with the following numbers superimposed on the screen: 20, 4, and 4. (This indicated that the same player continued to compete whereas the other two continued to cooperate.) Participants in the projection treatment once again saw and heard Player B accuse the others of lying just before the scores appeared on the screen. Participants were then instructed to stop the VCR and complete the final questionnaire.

At the end of the experiment, all participants were debriefed and asked not to discuss the details of the experiment with others. The debriefing consisted of probes for suspicion about the procedure (participants indicated very little suspicion about the experimental ruse and indeed frequently voiced their own theories of nonverbal cues and the detection of lying), a detailed presentation of the design and purpose of the study, and a discussion of how projection is often used in daily life.

INDEPENDENT VARIABLES

We manipulated three independent variables in the experiment.

Projection. The projection tactic was manipulated by having Player B accuse the other players of lying at the end of the 6th and 12th trials. Specifically, at the end of the 6th trial, Player B spoke the following, as he looked agitated and annoyed: “That means one of us was competing. Which one of you was lying?” The experimenter on the videotape quickly interrupted Player B and stated that the scores did indicate that someone was competing but that this could not be discussed now.

At the end of the 12th trial, Player B once again looked annoyed and said, “Someone was still lying!” He was then told by the experimenter on the videotape that scores could not be discussed at this time. Player B frowned and continued to look agitated in silence.

Participants in the no projection condition saw edited videotape with Player B’s accusations removed. It was identical in all other aspects.

Game players. The videotaped prisoner’s dilemma game was played by three “former participants” who were called Player A, B, or C. Player B was a man who delivered the projection tactic for half of the participants. Players A and C were both women.

Time of measurement. Participants completed the questionnaire about their perceptions of the players after the 6th and after the 12th trial.

DEPENDENT VARIABLE

The participants’ perceptions of Players A, B, and C, along with their opinions about the group, were assessed in a 12-item questionnaire administered after the 6^th and 12th game trials. The sole dependent measure in the experiment was a measure of each player’s culpability. This was assessed by six questions. Specifically, participants were asked, “How would you rate the playing strategy of player (with A, B, or C completing the stem)?” on a 9-point scale ranging from 1 (competitive) to 9 (cooperative). (This question was designed to indirectly assess participants’ perceptions of who was committing the misdeed of competing while they were indicating that they were cooperating.) Participants also were asked, “Do you think that player (A, B, or C) was honest when they displayed their cards or dishonest?” on a 9-point scale ranging from 1 (dishonest most of the time) to 9 (honest most of the time). (This question directly assessed the participants’ perceptions of who was to blame for dishonesty). These two measures were combined for each player to form a culpability index, with higher numbers meaning a more competitive and dishonest player. (The two items for each player were highly correlated with correlations between .52 and .87.)

The other six measures were filler items designed to disguise the purpose of the questionnaire. These items assessed perceptions of group cohesion, group playing style, group leadership, individual performance, and coalition formation (none of these filler items was affected by projection).

Results

Figure 1 presents the mean culpability (combined competitive and honesty measures) for each player as a function of the use of projection by Player B (collapsed across time of measurement). As can be seen, the use of projection as an interpersonal influence tactic greatly affected the culpability of each player as represented by a significant PlayerProjection interaction, F (2, 112) = 4.83, p< .02. Player B was seen as less culpable (M= 3.97) when he accused others of lying and failing to cooperate (compared to M= 5.34 in the control treatment). In contrast, both Players A and C were seen as more to blame when they were the target of Player B’s projection (M= 4.74 for Player A and M= 5.56 for Player C) relative to the control condition (M= 3.73 for Player A and M= 5.18 for Player C). The timing of measurement (whether perceptions were assessed after the 6th or 12th trials) did not interact with projection to influence culpability. ²

Discussion

Our first experiment clearly demonstrated the power of projection: When Player B accused others of misdeeds, he was exonerated at the expense of those he attacked. Thus, we appear to have successfully created an experimental analog of the attacks rendered by the Florentine aristocrats, Adolf Hitler, and Joseph McCarthy.

Interestingly, the time of measurement did not interact with projection, and thus, neither of our two preliminary hypotheses (that repetition of projection either strengthened the impact of the accusation or increased suspicion about the projectionist) was confirmed. There are multiple interpretations of this null result: (a) Both original hypotheses are wrong, (b) both are happening simultaneously and are canceling each other, or (c) the dose was not large enough to detect an effect.

EXPERIMENT 2: THE PROJECTION TACTIC IS EFFECTIVE EVEN WHEN SUSPICIONS ARE RAISED ABOUT THE PROJECTIONIST

The purpose of our second experiment was twofold. First, we conducted a replication of Experiment 1 to demonstrate the reliability of the projection effect. Second, we sought to examine a possible boundary condition for the projection effect: suspicion of the motives of the projectionist. In everyday life, demagogues who use projection are often viewed as having suspect purposes.

It is reasonable to assume that such suspicions would limit the effectiveness of the technique. When an observer is suspicious of a projectionist, he or she may scrutinize more carefully what is said and thus be in a better position to “catch the scoundrel” in a lie. Indeed, in such instances, an observer should be biased against the projectionist and thus question the validity of any accusation.

To meet our two goals, we replicated Experiment 1 but added an additional factor to the design—raising suspicions about the motives of the projectionist. This was accomplished by having participants read a set of background questionnaires that had been ostensibly completed by the game players. The information on these questionnaires varied to make Player B (the projectionist) appear either highly competitive or eager to win at all costs or of average competitiveness on par with the rest of the players. Method

PARTICIPANTS AND DESIGN

Participants included 43 University of California undergraduates who were randomly assigned in a 2 (projection: accusation vs. no accusation)2 (suspicion: low vs. high suspicions raised about the motives of the projectionist) 3 (game player: A, B, and C)2 (time of measurement: after the first accusation and after the second accusation) design (with game player and time of measurement serving as within-subject factors). The students received credit in their introductory psychology class for their participation.

PROCEDURE

We used the same procedures employed in Experiment 1—the same cover story, videotapes, and independent and dependent variables—with two exceptions. (In Experiment 2, the correlations between the competitive and honesty measures for each player ranged from .68 to .96 and were once again combined to form a culpability measure for each player.)

First, to create suspicions concerning the motives of the projectionist, each participant received “background questionnaires” that were supposedly completed by the game players the participant would be watching in the upcoming video. Participants received this information before viewing the videotape. Specifically, participants were told that the players had completed these questionnaires just before they had played the game to assess players’ expectations about the game.

Participants also were told that we were interested in seeing if they could predict players’ strategies (cooperative/ competitive) using this information. A separate background questionnaire was presented for each player.

The “questionnaires” were exactly alike in content and typeface except that each one had been filled out differently (in different inks and handwriting styles).

The questionnaire consisted of eight items (all answered on 5-point scales). Three items were critical for the manipulation. These questions asked, (a) “How important is winning this game to you?” (b) “How important is it for you to win lots of money at this game?” and (c) “Overall, do you plan to use a competitive or cooperative strategy?” In the low-suspicion condition, all three participants placed ratings of 1 or 2 for all three of these critical questions, thus indicating that winning was not all that important and that they planned to cooperate. In the high-suspicion condition, the same information was provided by Player A and Player C. However, Player B’s questionnaire now had ratings of 4 on the questions regarding the importance of winning and of winning money and a rating of 3 on the question about type of strategy. This indicated that Player B wanted to win and was considering adopting a competitive style. The remaining five items were filler material used to divert attention away from the manipulation. These items showed that all three players enjoyed working in groups, worked in groups often, liked to play games, expected the other players to be cooperative, and felt that losing did not bother them.

Second, participants also completed an additional questionnaire administered immediately after viewing the background questionnaires. The purpose of this measure was to assess the effectiveness of the suspicion manipulation. Three items on the questionnaire assessed participants’ expectations about the style of play each player would most likely use (competitive or cooperative). Specifically, the following question was asked about each player: “What type of strategy do you think Player ___ will adopt (with A, B, or C completing the question)?” Participants’ answered on a 5-point scale ranging from 1 (very cooperative) to 5 (very competitive). A fourth item assessed who participants thought would be the most competitive by asking participants the following question: “Who do you think is likely to accumulate the most points by the end of the game (circle one):

Player A, Player B, or Player C?” Two filler items were included to divert attention away from the purpose of these measures; these items asked about participants’ predictions about group cohesion and leadership. Completing this manipulation check just before viewing the videotape also served to make salient to the participant the motives of each player.

Results

MANIPULATION CHECK

Both sets of manipulation check measures indicated that the “player questionnaires” were effective in raising suspicions about Player B. When asked if Player B would be more likely to cooperate or compete, participants stated he would cooperate in the low-suspicion condition (M= 1.95) but compete in the high-suspicion condition (M= 3.59),F(1, 39) = 56.30,p< .01. The suspicion manipulation had no significant impact on perceptions of game strategy of the other two players, with means near the midpoint of the scale in both suspicion conditions for Player A (M= 2.46) and Player C (M= 2.42).

Similarly, in the low-suspicion treatment, only 19.1% of the participants thought that Player B would accumulate the most points compared to 47.6% for Player A and 33.3% for Player C. However, in the high-suspicion treatment, 59.1% thought Player B would be the high point winner compared to 18.2% for Player A and 22.7% for Player C, 2 (2,N= 43) = 7.65,p< .05.

PERCEPTIONS OF PLAYERS’ CULPABILITY

Figure 2 presents the mean culpability (combined competitiveness and honesty measures) for each player as a function of the use of projection by Player B and as a function of level of suspicion raised about Player B (collapsed across time of measurement). As can be seen, the use of projection as an interpersonal influence tactic once again affected the culpability of each player as rep-resented by a significant PlayerProjection interaction, F(2, 78) = 4.94,p< .01. This pattern of data occurred regardless of level of suspicion. In the low-suspicion condition, Player B was seen as having less blame (M= 3.50) when he accused others of lying and failing to cooperate (compared to M= 5.53 in the control treatment). A similar result occurred in the high-suspicion condition (M= 5.63 with projection vs. M= 7.00 with no projection). In contrast, both Players A and C were seen as having more culpability when they were the target of Player B’s projection (in the low-suspicion condition, M= 3.80 for Player A and M= 5.88 for Player C, and in the high-suspicion condition, M= 4.13 for Player A and M= 4.98 for Player C) relative to the no projection condition (in the low suspicion treatment, M= 3.37 for Player A and M= 4.05 for Player C, and in the high-suspicion condition, M= 3.70 for Player A and M= 4.30 for Player C). In other words, although Player B was seen as more culpable in the high-suspicion (M= 6.25) compared to the low-suspicion condition (M= 4.57), this did not mitigate or moderate the impact of his use of projection. (Suspicions about Player B did not affect the perceived culpability of Players A and C as indicated by a marginally significant PlayerSuspicion interaction (2, 78) = 2.51, p< .1.) As in the first experiment, the timing of measurement (whether blame was assessed after the 6th or 12th trials) did not interact with projection to influence culpability.

Discussion

Experiment 2 is noteworthy in at least two regards. First, we successfully replicated the projection effects obtained in Experiment 1: When Player B accused others of misdeeds, his own perceived honesty and cooperation improved while others were blamed for the misdeed. Second, raising suspicions about the motives of the projectionist did cause participants to perceive the accuser as more dishonest and less cooperative. However, these suspicions did not reduce the magnitude of the projection effect. The projection tactic appears to be effective even when the suspicion surrounding the projectionist is high.

Although we expected the suspicious motives of the projectionist to alert participants to the possibility of deception, our results parallel the findings of Toris and DePaulo (1984; see also Millar & Millar, 1997, for a recent replication and review of other research showing similar findings). In the Toris and DePaulo study, participants interviewed others who were either telling the truth or lying about their personality. Some of the participant interviewers were alerted to the possibility that deception may be occurring. Toris and DePaulo (1984) found that interviewers operating under suspicion (a) were not more accurate in detecting lies than those who were not told about the possibility of deception but (b) did tend to perceive more applicants as deceptive. Similarly, our participants who were alerted to the projectionist’s suspicious motives were not more accurate in debunking the projection tactic but did tend to perceive the projectionist as more culpable. Suspicion seems to lead to cynicism (the assumption of deception) as opposed to a careful appraisal of deceptive and misleading propaganda.

EXPERIMENT 3: PROJECTION IS EFFECTIVE IN A WAR SCENARIO WHERE OPPOSING EVIDENCE IS PRESENTED

In our first two experiments, we demonstrated that projection is an effective tactic for deflecting blame from oneself and placing it on another and that projection works even when suspicions are raised about the motives of the projectionist. Our third study had three objectives.

First, we sought to generalize the projection effect to another domain beyond the original prisoner’s dilemma game. In Experiment 3, participants read about a hostile conflict between two fictitious countries.

This conflict escalated until fighting broke out, resulting in heavy loss of human life. Before the fighting developed, one of the countries accused the other of preparing for an attack (the projection tactic). This scenario was designed to resemble the propaganda associated with events such as Hitler’s invasion of Eastern Europe and Iraq’s attack of Kuwait.

Second, given that just raising suspicions about the motives of the projectionist was not effective in reducing the impact of projection in Experiment 2, we took a more direct approach. In Experiment 3, we provided some of the participants with information from a neutral source (the news media), stating that the projectionist had initiated the attack. We hypothesized that such concrete information from a reliable source should serve to undermine the effectiveness of the projection tactic.

Finally, in Experiment 3, we began an exploratory investigation of the impact of projection on more global perceptions of the reputation and stature of the projectionist and target. In Experiments 1 and 2, we found that projection influenced participants’ perceptions on measures closely related to the content of the projection. In Experiment 3, we included a similar measure (blame for the attack) and also developed new measures to assess perceptions of each country’s general stature and character.

Method

PARTICIPANTS AND DESIGN

Participants included 60 University of California undergraduates who were randomly given one of four questionnaires in a 2 (projection: accusation vs. no accusation) x 2 (evidence of attack: none vs. opposite to projection) design.

PROCEDURE

Participants were approached in front of a college library and asked to complete a survey. The survey featured a 123-word description of a conflict between two countries: Klarvonia (the projectionist) and Pangeria (the target of the accusation). The scenario made the following points: (a) a large deposit of uranium was discovered on the border of Klarvonia and Pangeria; (b) both countries were very poor until this ore was discovered; (c) a dispute over mining rights had broken out between the two countries; (d) recently this dispute escalated and fighting broke out on the border; (e) there has been no confirmation of who initiated the attack but both sides suffered heavy losses with a total of 536 dead and 1,485 injured; and (f) a ceasefire has been declared along with a temporary restraint on mining. After reading this information, participants answered a set of 9 questions concerning their perceptions of the incident. After completing the questionnaire, the students were debriefed and thanked for their participation.

INDEPENDENT VARIABLES

We manipulated two independent variables in this experiment.

Projection. The projection tactic was manipulated by having Klarvonia hold a press conference where its leaders made the following statement accusing Pangeria of preparing for attack:

Pangeria has been gathering forces to launch an assault on our territory for months. Instead of sharing the uranium deposit equally, they wish to take the entire deposit for themselves. We believe that they plan to use violent means, in the near future, in an attempt to destroy our military installations. We do not want violence and would prefer peace. We will not be instigators. However, we will guard our borders and react to any action taken against us.

In the projection treatment, Klarvonia made this statement before the fighting broke out (after Point C in the scenario). The statement was not included in the control conditions.

Evidence of attack. In half of the questionnaires, we provided information that would lead to a conclusion opposite to that expressed by the projection statement. Specifically, participants read the following: “Rumors circulated in the news media that Klarvonia was responsible for the attack. Pangeria reports that their installation was destroyed....”In the evidence opposing projection condition, this statement was made just after the participants were informed that fighting had broken out on the border of the two countries (after Point D in the scenario). The statement was not included in the control conditions.

DEPENDENT VARIABLES

Participants’ perceptions of Klarvonia and Pangeria were measured in a nine-item questionnaire attached to the scenario. The primary dependent measure was an item that assessed blame for the attack. Eight additional items assessed the participants’ impressions of the overall stature of each nation.

Blame. Participants’ perceptions of who was to blame for the attack were assessed by asking, “Who do you think is responsible for initiating the attack?” Participants responded on a 9-point scale ranging from 1 (Klarvonia to blame) to 5 (equal blame) to 9 (Pangeria to blame). ³

Nation stature. The overall stature of each country was assessed by eight items (four per country). Specifically, participants were asked to respond on 9-point scales to the following items: “How justified was the nation of [Klarvonia or Pangeria] in the actions it took?” (Where 1 = unjustified and 9=justified). “How moral or immoral were the actions taken by the nation of [Klarvonia or Pangeria]?” (Where 1 = immoral and 9= moral). “How respectable is the nation of [Klarvonia or Pangeria]?” (Where 1 = respectable and 9=not respectable). “How aggressive were the actions taken by the nation of [Klarvonia or Pangeria]?” (Where 1 =aggressive and 9 =not aggressive).

Results

BLAME

Participants’ perceptions of who was to blame for the Klarvonia/Pangeria border attack were influenced by two factors. First, participants were more likely to blame Pangeria (the target of projection) when Klarvonia accused Pangeria of preparing for war (M= 5.23) compared to the no projection control condition (M= 4.50), F (1, 56) = 4.19, p< .05 (note that higher numbers indicate more Pangerian blame). This, of course, is the projection effect of deflecting blame to another by means of an accusation. Second, Klarvonia was more likely to be blamed for the attack when the news media offered evidence opposite to the projection tactic (M= 4.40) compared to when this statement was not made (M= 5.33), F(1, 56) = 6.79,p< .05. There was no interaction between projection and evidence against the accusation (F< 1).

Thus, projection was equally effective regardless of the information provided about the guilt of the projectionist.

NATION STATURE

None of the eight measures of national stature was affected by projection (or any interactions with projection). The only significant finding showed that participants felt that Klarvonia was less justified in its actions when the news media reported it initiated the attack (M= 3.8) compared to the control condition (M= 5.73), F (1, 56) = 15.55, p< .01.

Discussion

Experiment 3 obtained results consistent with the first two experiments. Once again, we found that accusing another of a misdeed diverted culpability away from the projectionist and toward the target of the accusation.

Furthermore, the effectiveness of projection was not moderated by information that contradicted the projectionist’s allegation: in this case, direct evidence from a neutral third party that refuted the content of the accusation. Consistent with Experiment 2 and the research of Toris and DePaulo (1984), direct evidence against the projectionist resulted in more blame but did not mitigate the effects of projection.

However, the effects of projection were limited to measures related to the specifics of the accusation (i.e., who was to blame for the attack) and did not affect more general measures of stature. (A similar lack of effects on more global measures will be obtained in Experiment 4.)

There are a number of possible reasons for the failure of projection to affect global measures, including (a) these measures are less sensitive than specific measures, (b) global measures are more likely to be influenced by a wider range of variables (implicit in the scenario) and not just projection, and (c) there may be some negative aspects to projection (such as being seen as a tattletale) that could counter the positive aspects of projection.

EXPERIMENT 4: PROJECTION IS EFFECTIVE (REGARDLESS OF TIMING) IN A CHEMISTRY TEST SCENARIO WHERE ALL CAN BE GUILTY

Experiment 4 attempted to accomplish four objectives. First, we sought to generalize the projection effect to another domain beyond those used in Experiments 1 through 3. In Experiment 4, participants read about a group of high school students studying for a chemistry test. A copy of the test was given to some of the students, who subsequently did quite well on the exam. One of the students who received a copy of the test and did well on it then accused other students of cheating (the projection tactic).

This scenario differs from those employed in the previous experiments by (a) using a different domain of

test-taking and academic performance, (b) guaranteeing high levels of suspicion by having the projectionist receive a copy of the stolen test and obtain the third highest grade on the exam, (c) including multiple targets of projection that could vary in level of suspicion (i.e., those students who received the test and those students in the class who did not), and (d) allowing the possibility that any number of people may or may not be guilty of the misdeed (i.e., theoretically everyone or no one could have cheated on the test) as opposed to the either/or structure of guilt (i.e., either the projectionist or the target of projection committed the misdeed) used in the first three experiments. We felt that these differences would reduce the magnitude of the projection effect because (a) our participant population would have more experience with test-taking and thus be in a better position to see through the projection scam; (b) there was clear incriminating evidence that the projectionist was guilty; (c) if projection worked, it would not be extended to innocent bystanders (those students in the class who did not receive a stolen test); and (d) participants were not required to render an either/or judgment but could see all as guilty (or not) of cheating.

Second, we manipulated the timing of projection so that it could be given before the teacher suspected that cheating had occurred or after the teacher graded the exam and announced to the class that she suspected that some had cheated. We expected that a projective accusation given after the teacher announced that cheating had occurred would be seen as more manipulative, self-justifying, and thus would be less effective.

Third, we continued the investigation begun in Experiment 3 on global perceptions of the reputation and character of the projectionist. Recall that in Experiment 3, these measures did not show the effects of projection that had been obtained with more specific measures. In Experiment 4, we included a measure closely related to the content of the projection (a likelihood of cheating measure) and also evaluative measures of the projectionist’s character.

Finally, we added a new measure to assess participants’ naive theory of the effectiveness of projection as an influence tactic. At the end of the experiment, we described a cheating student who used projection and then asked participants if this would be effective in shifting blame toward others.

Method

PARTICIPANTS AND DESIGN

Participants included 83 University of California and San Jose State University students who were randomly given one of three questionnaires: (a) a no projection control, (b) a projection before suspicion is aroused, and (c) a projection after suspicion is aroused. Participants completed the materials as part of a class assignment.

PROCEDURE

Participants were asked to read a 159-word scenario describing a high school chemistry test. The scenario made the following points: (a) the night before a chemistry test, a group of students gathered to study; (b) the group reviewed old tests provided by the teacher; (c) a student named Mike brought copies of the actual test; (d) Mike gave copies of the test to each student; (e) some students who received a copy of the test threw it away; (f) Alex, Carol, Debbie, Jack [the projectionist], and Sarah left with copies of the test; (g) after grading the test, the teacher was surprised that a number of students had scores that exceeded expectations; and (h) she thinks it possible that some of the students may have cheated and informs the class of her suspicions. After reading this information, participants answered a set of questions concerning their perceptions of the incident. After completing the questionnaire, participants were debriefed and thanked for their participation.

INDEPENDENT VARIABLE

We manipulated projection by having Jack (the projectionist) inform the teacher that Mike handed out copies of the test at the study session and that he suspects some of his classmates used it to cheat on the test. In the projection-before condition, Jack made this statement before the teacher had a chance to examine the test results (after Point F in the scenario), whereas in the projection-after condition, Jack made the statement after the teacher had graded the test and announced to the class that she suspects some students of cheating (after Point H). Jack did not say anything in the no projection control condition.

DEPENDENT VARIABLES

Participants’ perceptions of cheating on the chemistry test were measured by a 22-item questionnaire attached to the scenario. The primary dependent measures were 15 items that assessed the likelihood that each student in the class cheated on the test. Six additional items assessed the participants’ impressions of Jack’s (the projectionist) character. A final measure assessed the participants’ naive theory of how projection works. ⁴

Likelihood of cheating. Participants were asked to indicate the likelihood that each of 15 students in the class cheated on the test using a 9-point scale (where 1 = unlikely that student cheated and 9= likely that student cheated). Included on this list were Jack (the projectionist), Mike (who had stolen the test), and the other four students who left the study session with a test (Alex, Carol, Debbie, and Sarah) as well as a list of nine other students (Becca, Beth, Charlie, Cheryl, Dennis, Max, Pete, Richie, and Todd). In analyzing the data, we compared the likelihood of cheating measure for the projectionist (Jack) versus the test-sneakers (the mean likelihood rating for the five students who left with the test) versus classmates (the mean likelihood rating for the other nine students). This allowed us to look at the effects of projection on the projectionist as well as those under suspicion (test-sneakers) and those with little suspicion (classmates). For each student, we provided the score they had received on the test, which ranged between 60 and 98. To increase suspicion of cheating, Jack and Mike were given high scores of 93 and 98, respectively. For the other test-sneakers and classmates, an equal number of low and high test scores were given.

Character of the projectionist. The character of Jack (the projectionist) was assessed by six items. Specifically, participants were asked to rate Jack on a series of 9-point scales anchored by good/bad, trustworthy/untrustworthy, honest/dishonest, smart/dumb, overachiever/underachiever, and moral/immoral (with the first term in each pair = 1 and the second term = 9).

Naive theory of projection. To assess whether participants believe that projection is an effective tactic, the last item in the questionnaire asked,

Mike [the student who gave out the tests] fears that the teacher may discover the test was stolen. He hopes to avoid suspicion by telling her that another student has cheated on the test. If this was the only information she had, do you think this action would cause the teacher to be more or less suspicious of Mike?

Participants responded on a 9-point scale ranging from 1 (less suspicious of Mike) to 9 (more suspicious of Mike).

Results

LIKELIHOOD OF CHEATING

Figure 3 presents the mean likelihood of cheating for Jack, the test-sneakers, and classmates as a function of the use of projection by Jack. As can be seen, the use of projection once again was effective in deflecting guilt from the projectionist and toward others as represented by a significant StudentProjection interaction (4, 148) = 2.97,p< .05. Jack, the projectionist, was seen as highly likely to have cheated on the test in the no projection control treatment (M= 7.67). However, Jack was seen as less likely to cheat when he accused others of cheating both before the teacher graded the test (M= 6.50) and after the teacher graded the test (M= 6.32). In contrast, Jack’s accusation negatively affected perceptions of the rest of the class. The test-sneakers were seen as moderately likely to have cheated in the control condition (M= 4.89) and more likely to have cheated in the projection before (M= 5.90) and projection after (M= 5.19) treatments. Similarly, the rest of the class was seen as least likely to cheat in the control condition (M= 2.84). However, even they were not immune to the effects of projection and were seen as more likely to have cheated in both the projection before (M= 3.50) and projection after (M= 3.34) conditions.

CHARACTER OF THE PROJECTIONIST

Participants perceived Jack the projectionist to be significantly more honest when he used projection before the cheating was discovered (M= 5.13) and after cheating was discovered by the teacher (M= 4.71) than when he did not use projection (M= 6.52), F (2, 78) = 3.71,p< .05, with lower numbers indicating more honesty. However, no significant results were obtained for any of the more general measures of Jack’s character, including measures of good/bad, trustworthy/untrustworthy, moral/ immoral, smart/dumb, and overachiever/underachiever (all Fs < 1).

NAIVE THEORY OF PROJECTION

Finally, our measure of participants’ naive theory of the effectiveness of the projection tactic did not vary as a function of projection. However, the results did show that participants not only thought that projection would be an ineffective tactic but that it would likely backfire.

Given the scenario of Mike (the test stealer), who hopes to avoid suspicion by accusing others of cheating, participants indicated that such actions would cause the teacher to be more suspicious of Mike—a mean of 7.61 (on a 9-point scale). Looked at in another way, 15.8% of participants thought that projection would make the teacher less suspicious (a score of 1 to 4 on our naive theory measure), 7.3% thought it would be ineffective (a score of 5 or the midpoint of the scale), and 76.8% thought that projection would boomerang and lead to more suspicion of the projectionist (a score of 6 to 9).

Discussion

Experiment 4 replicates a now-familiar pattern: Projection leads to perceptions of reduced culpability for the projectionist and increased blame for those accused.

Raising suspicion about the act of projection (this time by making the accusation after suspicions of a misdeed had arisen) did nothing to mitigate the effects of projection. Furthermore, the projection effect occurred in a scenario with key features that should have reduced effectiveness (i.e., a domain familiar to the participants, inclusion of evidence incriminating the projectionist, innocent bystanders as target, and a judgment task that did not require an either/or guilt judgment). The effects of projection, however, do seem to be limited to the issues closely linked to the content of the accusation as opposed to general character and dispositional matters.

Experiment 4 did produce a rather interesting finding. When given a scenario of a likely cheater accusing others of cheating, the vast majority of participants believed that this action would bring more suspicion to the accuser. (Only 15.8% of the participants actually predicted the results of the experiment that projection would reduce suspicion.) What is remarkable about this finding is that not more than a minute or so earlier, two thirds of the participants had been confronted with a similar situation and behaved in the opposite manner. In other words, participants’ naive theory of projection as “likely to boomerang” was at odds with their behavior of reducing their suspicions of a likely cheater who had accused others of cheating. This pattern of data, of course, argues against a demand interpretation of the findings. It also provides a clue as to why projection is so effective in four experiments: Participants do not believe it will work and therefore take little precautions to detect and debunk projection.

GENERAL DISCUSSION

Summary of Results

This investigation found the following: First, across four experiments that varied in terms of domain (prisoner’s dilemma game, war scenario, academic test), format (audio-visual, written text), types of false accusations (lying and failing to cooperate, preparing for aggression, cheating), the timing of the accusation (before or after the misdeed was discovered by others), the type of person accused (fellow participant, enemy, classmate who likely cheated, innocent bystander), the number of people accused (one, two, or many), the number of people who could be guilty of the misdeed (from one to many), and the guilt of the accuser (from ambiguous to highly likely), we found consistent evidence for projection effects: Accusing others of a misdeed increased the blame placed on the target of the accusation and exonerated the accuser.

Second, raising suspicions about the projectionist’s motives and guilt did not eliminate or reduce the strength of the projection effects. Suspicion did create cynicism (more negative appraisals of the projectionist), but this did not lead to the detection of deceit inherent in projection.

Third, the effects of projection seem to be limited to measures specifically related to the accusation as opposed to more general measures of character and stature. As noted before, there are a number of possible reasons for this pattern of results, including global measures (a) are less sensitive than specific ones, (b) may be influenced by more than just projection, and (c) may pick up on some possible negative effects of projections (such as being seen as a tattletale). Future research will need to sort out these issues.

Fourth, our participants did not seem to think that projection would work to reduce the culpability of the projectionist even as the tactic was working on them. In fact, more than three quarters of the participants in Experiment 4 thought that projection would boomerang and bring more suspicion to the accuser.

Why is the Projection Tactic So Effective?

When we began this research, we had two goals: (a) to demonstrate the projection effect (which we accomplished) and (b) to provide a list of boundary conditions for the effect (which we failed to realize). In fact, we were quite surprised that our attempts to raise suspicions about the projectionist were so futile in eliminating the interpersonal effects of projection. We can suggest seven reasons why projection is such an effective influence tactic, both in our laboratory and in the social world.

1. People tend not to suspect that others may be attempting to deceive. In a review of lie detection research, DePaulo, Stone, and Lassiter (1985) draw an inescapable conclusion: Perceivers tend to give senders the benefit of doubt; people tend to believe what others tell them. In other words, the projectionist tends to be given the benefit of doubt, and thus steps are not taken that would be needed to uncover the deception.

2. People are reluctant to expose lies. In three provocative experiments, Taylor, Gittes, O’Neal, and Brown (1994) placed participants in a situation where they heard another person tell a lie. The lies could have had devastating consequences, either ruining a scientific experiment or putting others at risk (including the participant) of contacting a disease. Across all three experiments, only 18% of the participants were willing to expose the liar.

3. Factors that lead to lie detection are not present in our experiments or in many real-world applications of the projection tactic. Research on lie detection has found a number of factors that lead to the discovery of a lie, including (a) immediate inconsistencies in the story (Ekman, 1985), (b) presence of strong emotions that interfere with lie telling (Ekman, 1985), (c) nonverbal behavior such as blinking eyes and self-manipulating gestures such as rubbing and scratching (DePaulo et al., 1985), (d) verbal behavior such as high-pitched speech and grammatical errors (DePaulo et al., 1985), and (e) strange behavior and the violation of role expectations by the liar (Bond et al., 1992). These factors were not present in any of our four experiments and would not be present in many real-world applications of the projection tactic, especially those using the mass media where professional actors, publicity agents, unnamed sources, and leaks can be used to deliver the false accusations.

4. Factors that we did manipulate are unrelated to lie detection. In Experiments 2 through 4, we tried to raise suspicion about the projectionist through various means. As Toris and DePaulo (1984) found, suspicion can lead to cynicism (believing that everyone is lying) but does not necessarily lead to superior lie detection (see also Millar & Millar, 1997).

5. Projection initiates a number of cognitive responses favorable to the projectionist. It is useful to speculate what types of cognitive responses are elicited by a false accusation. We suspect that projection makes more likely a variety of cognitive responses, including (a) creating a more favorable contrast between the accused and projectionist, (b) demonstrating that the projectionist is concerned about moral behavior, and (c) diverting attention away from any flaws in the character and behavior of the projectionist; in other words, it is the accused and not the projectionist who is the focus of attention and investigation.

Such misdirection is a key ingredient in successful stage magic (a form of deception; see Schiffman, 1997) as well as successful lying in general. Of course, there may be other cognitive responses, and there is no reason to assume that these responses are mutually exclusive.

6. People may not be aware of the effectiveness of projection in redirecting blame. In Experiment 4, the vast majority of participants thought that projection would not work and would boomerang to increase the perceived guilt of the accuser. To the extent that participants feel the tactic will not work or are overconfident in their ability to detect it, then they will not take needed precautions to identify and defuse the effects of projection.

7. Once the misinformation of projection is accepted, it serves to guide future perceptions and beliefs that can then reinforce the original misinformation. Allport and Postman (1947) found that information such as a rumor is often changed and distorted to fit the receiver’s expectations. In review-ing two decades of recent memory research, Loftus (1992) concludes that misleading information can turn a lie into memory’s truth. In other words, the misinformation of projection can serve as a frame to encode and interpret new information, which in turns confirms the original misinformation. For example, in Experiments 1 and 2, projection may have focused the participants’ attention away from the projectionist and toward the other players, whose nonverbal and verbal behavior might then be scrutinized for clues as to which of them is lying. Any suspicious information is then taken as evidence for the validity of the projection (Bond & Fahey, 1987). In Experiments 3 and 4, direct evidence against the projectionist may have been discounted—the news report of the projectionist’s aggression is disregarded as unsubstantiated, and the projectionist’s high grade on the exam is seen as the mark of a good student—all in an attempt to fit new information with the “truth” of the false allegation. As more and more information is assimilated as consistent with the false allegation, it becomes harder to detect and refute the projection. Furthermore, should a lie start to come to light, a projectionist can offer excuses and justifications and thus avoid any negative consequences of detection (Shapiro, 1991).

What Can Be Done About Projection as Propaganda?

One of the 10 Commandments that God ostensibly gave to Moses states, “You shall not bear false witness against your neighbor” (Exodus 20:16). In other words, according to the Book of Exodus, projection is, for the most part, a sin. Common sense notions of morality agree. Projection often involves egoistic motives (a liar attempts to gain at the expense of others) and is about matters of importance—two factors identified by Backbier, Hoogstraten, and Terwogt-Kouwenhoven (1997) that lead others to deem a lie as socially unacceptable.

Not much analysis is needed to understand why the widespread use of projection can have devastating social consequences. Imagine a society where false accusations are rampant or an organization where projection passes for administrative process. (Perhaps you don’t have to.)

In such situations, scoundrels are exonerated and then praised, innocents have their reputations tarnished and then destroyed (if they are lucky), and eventually fewer and fewer communications can be trusted (Sagarin, Rhoads, & Cialdini, 1998). What can be done about the negative consequences of projection?

One approach to countering projection as propaganda is to stop it at the source. Many societies have developed strong norms against bearing false witness and strong sanctions to support those norms. For example, the ancient Israelites were required to atone for their sin of false witness. In ancient Greece, those who brought false accusations in a court of law could suffer the same sanctions—banishment, death, forfeiture of property—as the target of those accusations would have suffered if the false witness had been believed. In 17thcentury Britain, those who gossiped and lied about their neighbors were required to don the brank or scold’s bridle, an iron cage locked on the head and equipped with a spiked tongue depressor that prevented the wearer from talking. Among the West African Ashanti, nasty rumors are punished by cutting off the gossiper’s lips, and the Seminole Indians treat talking bad about someone as if it was stealing (Levin & Arluke, 1987).

Such norms and sanctions are a far cry from U.S. defamation law established in New York Times v. Sullivan, which places a considerable burden of proof on the falsely accused to obtain a redress of damage (see Schauer, 1980).

However, strong norms and sanctions against projection are of little value if the deceit cannot be detected. As Ekman (1985) points out, “Lie checking isn’t a simple task, quickly done” (p. 240). However, the chances of catching a deceptive projectionist can be increased by examining the underlying motivation of the source of the information and by investigating the consistency of the story (see Pratkanis & Aronson, 2001). This can be accomplished by asking such questions as “Why is this person telling me this? What do they have to gain? What is the evidence for this statement? What are the arguments for the other side?” In the absence of definitive answers to these questions, a wise course would be to postpone judgment.

Of course, these questions for detecting deceptive projection are of little value if they are never asked. And they will not be asked as long as there is doubt, as with our participants in Experiment 4, about the effectiveness of the projection tactic for shifting blame away from the projectionist and toward the accused—a doubt this research should lay to rest.

NOTES

1. In Experiments 1 and 2 probing at debriefing revealed that four participants were suspicious of the validity of the tapes; we did not include their data in the analysis.

2. Figures 1 and 2 also show differences in the three players (on culpability) in the no projection treatment of Experiments 1 and 2. This was a function of natural variation in the filming of the tape. For example, Player C appeared shorter on the film and tended to look down a lot, thus appearing nervous. Such variations led to the differential perceptions. After filming, we thought this variation was an interesting aspect of the design because it further demonstrated the robustness of the effects of projection.

3. We also included an ancillary measure asking participants to assign 100 United Nations investigators to assess the problems between Klarvonia and Pangeria but will not consider this measure further due to interpretational ambiguities.

REFERENCES

Allport, G. W., & Postman, L. (1947). The psychology of rumor. New York: Henry Holt.

Backbier, E., Hoogstraten, J., & Terwogt-Kouwenhoven, K. M. (1997). Situational determinants of the acceptability of telling lies. Journal of Applied Social Psychology, 27, 1048-1062.

Bond, C. F., & Fahey, W. E. (1987). False suspicion and the misperception of deceit. British Journal of Social Psychology, 26, 41-46.

Bond, C. F., Omar, A., Pitre, U., Lashley, B. R., Skaggs, L. M., & Kirk, C. T. (1992). Fishy-looking liars: Deception judgment from expectancy violation. Journal of Personality and Social Psychology, 63, 969-977.

de Grazia, S. (1994). Machiavelli in hell. New York: Vintage.

DePaulo, B. M., Stone, J. I., & Lassiter, G. D. (1985). Deceiving and detecting deceit. In B. R. Schlenker (Ed.), The self and social life (pp. 323-370). New York: McGraw-Hill.

Ekman, P. (1985). Telling lies. New York: Norton.

Fine, G. A. (1992). Manufacturing tales. Knoxville: University of Tennessee Press.

Freeman, E. (1940). Conquering the man in the street. New York: Vanguard.

Levin, J., & Arluke, A. (1987).Gossip: The inside scoop. New York: Plenum.

Loftus, E. F. (1992). When a lie becomes memory’s truth: Memory distortion after exposure to misinformation. Current Directions in Psychological Science, 1, 121-123.

Millar, M. G., & Millar, K. (1997). Effects of situational variables on judgments about deception and detection accuracy. Basic and Applied Social Psychology, 19, 401-410.

Newman, L. S., Duff, K. J., & Baumeister, R. F. (1997). A new look at defensive projection: Thought suppression, accessibility, and biased person perception. Journal of Personality and Social Psychology, 72, 980-1001.

Pratkanis, A. R., & Aronson, E. (2001). Age of propaganda: The everyday use and abuse of persuasion (Rev. ed.). New York: Freeman.

Rovere, R. H. (1959). Senator Joe McCarthy. New York: Harcourt Brace Jovanovich.

Sagarin, B. J., Rhoads, K. L., & Cialdini, R. B. (1998). Deceiver’s distrust: Denigration as a consequence of undiscovered deception. Personality and Social Psychology Bulletin, 24, 1167-1176.

Schauer, F. (1980). Social foundations of the law of defamation: A comparative analysis. Journal of Media Law and Practice, 1, 3-23.

Schiffman, N. (1997).Abracadabra! Amherst, NY: Prometheus Books.

Shapiro, D. L. (1991). The effects of explanation on negative reactions to deceit. Administrative Science Quarterly, 36, 614-630.

Taylor, L., Gittes, M., O’Neal, E. C., & Brown, S. (1994). The reluctance to expose dangerous lies. Journal of Applied Social Psychology, 24, 301-315.

Toris, C., & DePaulo, B. M. (1984). Effects of actual deception and suspiciousness of deception on interpersonal perceptions. Journal of Personality and Social Psychology, 47, 1063-1073.

Wegner, D. M., Wenzalaff, R., Kerker, R. M., & Beattie, A. (1981). Incrimination through innuendo: Can media questions become public answers? Journal of Personality and Social Psychology, 40, 822-832.