PART II

EMPIRICAL PAPERS

PAPER 7

Ally maintenance reputation gossip about celebrities:

we share good and bad gossips about celebrities we like, and do not share gossip about celebrities we dislike

Abstract

The scope of this study is to test whether we feel happy to hear good gossip about celebrities we like and bad gossip about celebrities we dislike, and whether we feel distressed when hearing bad gossip about celebrities we like and good gossip about celebrities we dislike. In addition I also investigate if good gossip about liked celebrities and bad gossip about disliked celebrities is more easily passed on to others than bad gossip about liked celebrities and good gossip about disliked celebrities. To test these predictions I set up a list of 9 good gossip and 9 bad gossip stories. A control group of 40 respondents rated these gossip stories on a 1-7 Likert scale and confirmed that the bad/good connotations I had attributed to the gossip stories. Next I presented these good and bad gossips to 942 Flemish (Belgian) and American respondents. I used 18 well known celebrities as gossipees.

Results show that bad gossip about celebrities is not likely to be shared with others. In contradiction to my predictions those respondents who disliked the gossipees of bad gossips did not have a stronger tendency to share these gossips with others. Emotional responses of the respondents did however indicate that those who liked the gossipees of bad gossips felt more distressed to hear these bad gossips than the respondents who disliked the gossipees. For the good gossip stories however, predictions can be confirmed. Those who like the gossipees of good gossips feel happier to hear this and are more willing to share this gossip information with others than those who dislike the gossipees of good gossips. Overall it follows from this study, that respondents who like celebrities share more gossip about them with others than respondents who dislike celebrities.

1 Introduction

In this paper I focus on one of the different kinds of gossip I differentiated for in chapter four of the theoretical part of this dissertation: Ally Maintenance RG. I explained in that chapter how this form of gossip functions to maintain structural bonds with real life social network members. We gossip good about our friends, bad about our rivals and foes and remain neutral about those we regard as neutral social network members because these strategies in the end rebound to private status increases.

In chapter seven of this theoretical part of this dissertation I then explained how all different kinds of gossip, both Strategy Learning Gossip and Reputation Gossip make sense for celebrities as gossipees for some of us. I here focus on the use of Ally Maintenance RG with celebrities as gossipees. To start I will briefly repeat the theoretical background behind my hypotheses first.

1.1 Celebrity Gossip: learning or parasocial interacting

Celebrity Gossip is part of Media Gossip, which embodies gossip stories about both celebrities and public unknowns who appear in the media for a short term period. Celebrities are media characters who turn up in the media for long term periods. Their faces pop up in magazines, on television and on billboards along the roads of Western societies. In chapter 7 and in previous papers I have outlined that two hypotheses explain our interest in gossip about celebrities.

According to the so called Learning Hypothesis gossip stories about celebrities appeal to everyone because we can learn from their behaviors. Gaining experience through gossip is a fast and frugal means to increase our knowledge about fitness-relavant strategies (e.g. how to stay healthy, how to attract mates and so on). We constantly learn, by own experience and experiences of others, what are good strategies to use in the future and which strategies we better not use, because of their costly outcome. Strategy Learning Gossip, which is gossip about strategies others have tested, is valuable regardless whether the gossipee is known to the gossipers or not, because the focus lies on the transmitted behavior information. Unknowns can become gossipees of SLG if we can learn about the outcome of their tested strategies. Celebrities are unknowns in the sense that we on average never establish real life encounters with these people. However, they have a special feature, which is there high status. As I have explained more in detail in chapter four and chapter five, their high status implies that they are not only subject to SLG with a clear outcome of the gossiped about strategy, but also of SLG about strategies that do not have a clear positive or clear negative outcome. This is because we want to copy these higher status individuals, because mimicking higher status others can increase your own status (Henrich & Gil-White, 2001). And as Boyd and Richerson (1985) explained a fast and frugal strategy to copy the successful strategies of higher status others is to copy the overall behavior pattern of higher status others (General Copying Bias).

The Parasocial Hypothesis adds to this that some of us value gossip about celebrities because they regard celebrities as being part of their social network. Because we somehow encounter celebrities through media images on a daily basis, some people start regarding some celebrities as social network members. The difference with real life social network members is that interactions with celebrities are one-way or parasocial (Caughey, 1984 Horton and Wohl, 1956; Rubin, Perse, and Powell, 1985). Celebrities reveal their private lives (willing or unwilling) and we show emotional responses to this. But we do not reveal our private lives to them (although some would like to), and they do not get emotional over us. Still, because of these one-way interactions with celebrities some people are interested in gossip about celebrities for the exact same reasons as we are all interested in gossip about real life social network members. Because we want to learn about specific others and we want to manipulate their reputations. As Barkow (1992) nicely framed it:

A possible answer is that the mass media may activate the psychological mechanisms that evolved in response to selection for the acquisition of social information. […] We see them in our bedrooms, we here their voices when we dine: If this hypothesis is correct, how are we not to perceive them as our kin, our friends, perhaps even our rivals? As a result, we automatically seek information about their physical health, about changes in their relative standing, and above all about their sexual relationships.”(Barkow, 1992: 629-630)

1.2 In-group out-group in evolutionary perspective

Krebs and Denton (1997) stressed the importance of judging others in terms of friends or enemies. Small cooperative groups competed with each other in our evolutionary past (Krebs & Denton, 1997; Lewin, 1993; Tooby & Devore, 1987). Increasing the status of the group with whom an individual is associated rebounds to the fitness of that individual, and is therefore an adaptive strategy. Therefore we tend to put in-group members in a positive perspective and out-group members in a negative perspective, they say. And a simple means to do this is by using good and bad gossip.

1.3 Good gossip and bad gossip in interpersonal interactions

McAndrew and Milenkovic (2002) argued that from an evolutionary perspective, individuals should be eager to spread good gossip about allies and bad gossip about non-allies. Negative information lowers the reputation of other individuals, and is therefore an ideal strategy to tackle higher status people or rivals, they say. Likewise, by elevating the status of allies one can relatively increase his or her own status, they say.

I label this specific functioning of gossip Ally Maintenance Reputation Gossip. Ally Maintenance RG functions to maintain social structures. By praising our allies we not only increase our relative status, but also strengthen the bonds with our allies, by favoring them with an increase of their status as well. Likewise by slandering our rivals and foes we relatively increase our own status, and we decrease their status, which enforces our negative bond with them.

Investigating the use of Ally Maintenance Reputation Gossip McAndrew and Milenkovic (2002) surveyed 83 respondents (39 males, 44 females) ranging in age from 17 to 22 about how likely they would share good/bad information about friends, relatives, acquaintances, professors, or strangers. They predicted that good gossip will be shared about allies as gossipees and bad gossip will be shared about non-allies as gossipees. In proof of this, their results indicate that individuals more likely share negative gossip about potential adversaries (strangers and powerful others) and do not share negative information about allies (friends and relatives). Good gossip was only shared when gossipees were allies and not when the subjects were non-allies:

“People actively seek information about others that will be most useful in social competition. We seek exploitable, damaging information about high-status people and nonallies; we actively disperse status-enhancing information about our allies; and we keep a very watchful eye on our friends.” (McAndrew & Milenkovic, 2002: 17)

However, what I comment is that McAndrew and Milenkovic’s (2002) research only supports part of the Ally Maintenance RG predictions. They have shown that we are more prone to spread good news about allies, but their results about ‘non-allies’ should be taken with some care. Professors are not necessarily non-allies. Some students might regard some professors (they look up at) as allies. Second, strangers are not necessarily non-allies. If we do not know them yet, we might have not yet classified them as friends or foes. As a third comment I do not go along with their statement that we will automatically gossip bad about higher status others. Whether we will use good or bad gossip about higher status others depends on whether we regard this higher status individual as a friend or as a foe. We will not slander a higher status person who is allied to us because we can bask in the glory of being connected to this person, and we will praise this person instead.

I further comment that the ‘good’ and ‘bad’ information they used should be tested. What they considered as ‘bad’ was for instance drug abuse, but among students (their research population) using drugs might not necessarily have such a negative connotation as it has among adults.

2 Hypotheses

From the above outlined theoretical background I now put forward some specific hypotheses to test whether Ally Maintenance RG is also used to praise our one-way friends and slander our one-way foes. Regarding celebrities as one-way friends and one-way foes I expect that respondents will report positive or negative emotions to receiving good or bad gossip about a celebrity. More in detail, I predict the following:

Hypothesis 1a:

Respondents who strongly dislike a celebrity will report positive emotions when hearing bad gossip about this celebrity.

Hypothesis 1b:

Respondents who strongly dislike a celebrity will report negative emotions when hearing good gossip about this celebrity.

Hypothesis 1c:

Respondents who strongly like a celebrity will report negative emotions when hearing bad gossip about this celebrity.

Hypothesis 1d:

Respondents who strongly like a celebrity will report positive emotions when hearing good gossip about this celebrity.

Similar as McAndrew and Milenkovic’s (2002) predictions about the sharing/not sharing of good/bad gossip about real life friends and foes, I make following predictions about the sharing of good/bad gossip about celebrities as one-way friends and one-way foes:

Hypothesis 2a:

Respondents who strongly dislike a celebrity will share bad gossip about this celebrity with others.

Hypothesis 2b:

Respondents who strongly dislike a celebrity will not share good gossip about this celebrity with others.

Hypothesis 2c:

Respondents who strongly like a celebrity will not share bad gossip about this celebrity with others.

Hypothesis 2d:

Respondents who strongly like a celebrity will share good gossip about this celebrity with others.

3 Methodology

To test this I distributed a survey to 104 American University of California Santa Barbara (UCSB) students and 838 Flemish (Belgian) people of different age groups to test these hypotheses. This survey was part of a bigger survey about interest in media gossip in general, out of which I have discussed results already in the two previous papers.

I first asked my respondents to rate how much they liked/disliked a list of 18 celebrities that are all well known in Western societies. Next, about each of these 18 celebrities I either presented them a gossip story that increases the reputation of the gossipee (good gossip) or decreases the reputation of the gossipee (bad gossip). Working with two versions of the survey (A and B), I presented a ‘good gossip’ and a ‘bad gossip’ about every celebrity listed in the list of 18 well-known celebrities. For an overview of the gossip stories I used, as well as the celebrities I used as gossipees, I refer to table 7.1.

For survey A I received 52 complete questionnaires from American UCSB students and 378 complete questionnaires from Belgian (Flemish) respondents. For survey B I received 52 complete questionnaires from 52 American UCSB students and 760 complete questionnaires from Belgian (Flemish respondents). Both groups of respondents (survey A and survey B) do not differ significantly on average age and sex ratios (for descriptions see paper 5).

To test whether the presented gossip stories are really as good or as bad as I thought they were, I presented them to a control group. All names of the gossipees were removed; I simply asked 40 graduate philosophy students to rate to what extent they considered the uses traits/behaviors as appropriate, good behavior or inappropriate, bad behavior. All 40 students rated all 18 gossip stories on a likert-scale, ranging from 1 (bad behavior) to 7 (good behavior). Results will be discussed in the next section.

After the respondents read the good/bad gossip about a celebrity they were asked two questions. I first asked them to rate on a Likert scale from 1-7 how they felt hearing this information: distressed (1) or glad to hear this (7). I herewith tested if they showed positive or negative emotions towards the information. The second question concerned their tendency to share this gossip story with other (unidentified) people. Again they had to rate on a Likert scale from 1-7 how likely they would share this gossip with other people, with (1) not at all and (7) very much likely.

Table 7.1. Used good/bad gossip stories about media gossipees

4 Results

In what follows I will first of all discuss the results of the control experiment to test how bad and how good my presented good/bad gossip statements were. I will then give an overview of how liked and disliked the involved celebrities were rated by all 942 respondents. Last, I turn to the core of this study by giving an overview of all results for all used good and bad gossip stories, from which I will draw some tentative conclusions at the end of this result section.

4.1 How good is good and how bad is bad

The control group of N=40 students consisted of 57.5% males and 42.5% females, aged between 20 and 35, with an average of M_age=23.18 (S.D.= 3.28). They all rated all 18 stories on a 1-7 Likert Scale. All nine gossip stories I had set up as ‘bad gossips’ were indeed rated as negative and all nine ‘good gossips’ were rated positive. All good gossips have average scores higher than 5.00 on a 1-7 Likert scale, and all bad gossips have an average score lower than 3.00 on a 1-7 Likert scale. For an overview of all average scores of the 18 gossip stories I refer to table 7.2.

Table 7.2 Ratings of the good and bad gossip stories by a control group of N=40 students

4.2 Least liked and best like celebrities

I averaged the mean score of all 942 respondents (both American and Belgian) of their liked/disliked ratings for each celebrity I listed in the list of 18 well-known celebrities. An overview can be found in table 7.3. Overall Mother Theresa came out as the best liked person, followed by Princess Diana, Madonna, J.F. Kennedy and Elvis Presley. Five least liked persons of all 18 celebrities I presented my respondents were Adolph Hitler, Saddam Hussein, President Bush Jr. and Fidel Castro.

Table 7.3. Ranking and liked/disliked ratings of all N=942 respondents for all 18 well known celebrities

4.3 Good gossip about liked and bad gossip about disliked celebrities

To investigate whether we enjoy hearing good gossip about celebrities we like and bad gossip about celebrities we dislike, and whether we are more prone to share this with others than bad gossip about celebrities we like and good gossip about celebrities we dislike, I only took into account those respondents who showed a strong like/dislike attitude towards the listed celebrities. This is, I only looked at those who indicated to strongly dislike, by filling out 1 or 2 on the 1-7 likert-scale and strongly like, by filling out 6 or 7 on the 1-7 likert scale. All answers from 3-5 on the scale I consider being neutral and were left out of the analyses.

For all analyses I used a T-test for Equality of Means for Independent Groups. I will discuss both the results on whether respondents report if they will share or not share these gossip stories with others and how they feel when learning the transmitted good/bad gossip about a specific celebrity.

4.3.1 Bad gossips about liked and disliked celebrities

I will start with the results for how respondents reacted on the ‘worst gossip’, the bad gossip story that was rated by the control group as most inappropriate behavior. This concerns the gossip story about acting aggressive towards your personnel and slapping them in the face (gossip 5 in table 7.1). This story got an average rating of 1.05 on a 1-7 Likert scale (see table 7.2). Gossipees of this story were Princess Diana (survey A) and Grace Kelly (survey B). For both I found similar results (see table 7.4). Both those respondents of survey A, who dislike and like Diana and those respondents from survey B, who dislike and like Grace Kelly are not happy to hear this information. For both surveys the respondents who like either Diana or Grace Kelly are significantly (p<.01) more distressed than those who do not like either Diana or Grace Kelly. This is not completely in line with what I expected. I had expected those who dislike Diana and Grace to like this kind of bad gossip about her, still results indicate they do not. The significance difference with those who do like Diana and Grace Kelly however indicates a tendency of the expected (hypotheses 1a and 1c) difference between both groups.

Table 7.4. Responses to bad gossip about Princess Diana and Grace Kelly

However, when looking at the results for sharing this bad gossip with other people I cannot confirm my predictions 2a and 2c at all. I had expected those who dislike Diana and Grace Kelly to share this bad gossip about them with others, while those who like Diana and Grace Kelly not to share this bad gossip with others. The results for both (see table 7.4) show however that both groups of respondents are not eager to share this bad gossip with others, and that those who dislike Diana even significantly (p<.05) are less prone to share the bad gossip about her than those who like Diana. For Grace Kelly as well I found an indication that those who like her would be more willing to share this bad gossip about her than those who dislike her. However the difference is not significant.

Turning to the next bad gossip story (average rating 1.15 see table 7.2), which deals with Saddam Hussein (survey A) and Fidel Castro (survey B) who beat up their wife and children (story 4 in table 7.1), I found again that on average both those who like and dislike Castro and Hussein feel distressed to hear this information and are not very likely to pass this on to others. Again my hypotheses 1a and 1c predict that those who dislike Castro and Hussein should on average be less distressed and more eager to share this information with others than those who like Castro and Hussein. When looking at the results for emotional responses to this gossip story (see table 7.5) hardly any difference occurs in the emotional responses of the respondents who like or dislike Castro. Both groups feel quite distressed to hear this information. The results for Hussein have to be interpreted with some care, since only 4 respondents indicated a positive attitude towards Hussein. This results in a small number of respondents in the “Like Hussein” group. Still, comparing their results to those of the 376 respondents who strongly dislike Hussein, its shows (see table 7.5) that those who like Hussein significantly (p<.05) feel less distressed to hear this bad gossip than those who dislike him. Which is opposite to what I predicted, but the results are not significant and the number of respondents is too low to draw conclusions.

Table 7.5. Responses to bad gossip about Fidel Castro and Saddam Hussein

The results for sharing this bad gossip with others show some proof of my predictions for Castro as gossipee. Those who dislike Castro seem a little (3.44 vs. 2.81) more likely to pass on this bad gossip to others. Still the difference is not very significant (p<.1). The results for Hussein as gossipee again contradict my predictions 2a and 2c. Both those who like and dislike Hussein are not very eager to pass on this information to others (3.75 and 3.59). I remark again, however, that the group of respondents that like Hussein is too small (N=4) to draw strong conclusions, and the results are again not significant.

The third most negative gossip story, according to the ratings of the control group was the gossip about a gossipee who made racist remarks and denigrated people with different skin color. This gossip (3 in table 7.1) got an average rating of 1.20 and a 1-7 Likert scale. Gossipees of this bad gossip were Marilyn Monroe (survey A) and Prince Charles (survey B). When looking at the results for the emotional responses of those who like or dislike Charles and Monroe, I notice for both that the respondents who dislike the gossipee feel a little less distressed than those who do like the gossipee of this bad gossip, however the differences are non significant (see table 7.6).

Table 7.6. Responses to bad gossip about Marilyn Monroe and Prince Charles

In line with the two bad gossip stories I just discussed, I here as well notice that respondents who like Monroe or Charles are more eager to share this bad gossip with others than those respondents who do not like Monroe or Charles. For Marilyn Monroe this difference is highly significant (p<.001). Those respondents who like her are not very willing to share the news with others (3.83), but still more than those who do not like her (1.88). For Prince Charles both means are not that much different (not significant). Those respondents who like Charles are moderately willing to share this news with others (4.00), and those who dislike Charles show an average similar tendency (3.41).

The fourth most bad gossip story, according to the control group concerned the gossip about a gossipee who was also known for his/her racist remarks and had even received a fine for his/her inappropriate behavior (gossip 3 in table 7.1, with an average rating of 1.20; see table 7.2). Gossipees here were Brigitte Bardot (survey A) and Ronald Reagan (survey B). Looking at the results (see table 7.7), again for both Bardot or Reagan as gossipee respondents who dislike the gossipee indicate to feel less distressed to hear this bad gossip about them, than those respondents who like Bardot or Reagan. However these differences are non significant.

Table 7.7. Responses to bad gossip about Brigitte Bardot and Ronald Reagan

When looking at the results for sharing this bad gossip with others, I again remark an opposite trend as what I predicted in hypotheses 2a and 2c. Respondents who like either Bardot or Reagan are more willing to pass on this bad gossip than those respondents who do not like Bardot or Reagan. Still, also the differences between both groups here are non significant. There is a trend in the opposite direction (compared to the predicted direction), but this trend is not significant, so no conclusion can be drawn from this example.

Turning to the fifth most negative gossip story, I present the results for the gossip story that spreads the news about a gossipee who had fired his/her butler because this butler had spilled red wine on a white carpet gossip 18 in table 7.1). The control group again rated this as a negative gossip story (mean score= 1.55). Gossipees of this story were J.F. Kennedy (survey A) and Adolph Hitler (survey B). Similar as with Saddam Hussein, I hardly have any respondents who reported to really like Adolph Hitler. The group of “Like Hitler” consists out of only 3 respondents; the results for this group have to be taken with much care again. The group of people who report to really dislike Hitler is very big, with N=464 people. Similar to the above discussed results I again notice that all respondents feel rather upset hearing this bad gossip. Still there is a tendency again that those who dislike Kennedy and Hitler feel less distressed than those respondents who report to like Kennedy or Hitler. For Kennedy as gossipee the difference between the emotional response of those who like him is even significantly (p<.01) stronger negative than from those who report to dislike Kennedy (3.62 vs. 3.08).

Table 7.8. Responses to bad gossip about J.F. Kennedy and Adolph Hitler

Again, when I look at the tendency of the respondents to share this bad gossip with others I notice an opposite trend than what I predicted in hypotheses 2a and 2c. For both Kennedy and Hitler as gossipee of this bad gossip story, the respondents who report to like Kennedy or Hitler are more eager to share this gossip with other. I remark however, that similar to previous results the tendencies to share this bad gossip with others is not very big. For Kennedy as gossipee the mean score is 3.27, which still indicates a tendency to not share this information with others. For Hitler as gossipee I noticed a positive mean score of 5.67, which indicates they probably would share this gossip with others, however I mention again that this group only consists out of 3 respondents, and only indications and no conclusions can be drawn from this.

The sixth rated bad gossip story is story 7 in table 7.1 about Madonna (survey A) and Michael Jackson (survey B) who used to tease children at school. This was rated negative with an average score of 1.80 on a 1-7 Likert scale. The results for the emotional responses of the respondents (see table 7.9) show no significant differences between those from survey A who like or dislike Madonna and those respondents of survey B who like or dislike Michael Jackson. All respondents report to feel distressed hearing this gossip.

Table 7.9. Responses to bad gossip about Madonna and Michael Jackson

For their tendency to share this gossip with others, all respondents indicate not to be willing to share this bad gossip with others. Still for Madonna as gossipee, those respondents of survey A who like her have a highly significant (p<.001) stronger tendency to share this with others than those respondents of survey A who do not like her. This is again an opposite trend of what I predicted in hypotheses 2a and 2c.

The next bad gossip is the one about President Bush Jr. (survey A) and Arnold Schwarzenegger (survey B) being ex-cocaine addicts (story 2 in table 7.1). The N=40 control group respondents gave this gossip an average rating of 2.05 on a 1-7 Likert scale. Results for the emotional responses to this bad gossip show significant (p<.01 for Bush Jr. as gossipee and p<.001 for Schwarzenegger as gossipee) less distressed responses from the respondents of surevy A who dislike President Bush Jr. and the respondents of survey B who dislike Arnold Schwarzenegger than the emotional responses of the respondents of survey A who like President Bush Jr. and the respondents of survey B who like Arnold Schwarzenegger. This confirms my predictions 1a and 1c.

Table 7.10. Responses to bad gossip about President Bush Jr. and Arnold Schwarzenegger

The results for the tendencies to share this bad gossip with others are in line with the other results and contradict my predictions 2a and 2c. None of the respondents is eager to share this gossip with other. Those who dislike Bush Jr. or Schwarzenegger are not more willing to spread around this bad gossip about their one-way foes. The results here even show no difference at all between the tendency to share this bad gossip of the respondents who like or dislike the gossipees.

The eight most negative ranked gossip story is about Mother Theresa (survey A) and Bill Clinton (survey B) who can be real nasty people yelling at others for no reason (gossip 13 in table 7.1). The average rating for this bad gossip from the control group is 2.45 on a 1-7 Likert scale. For the respondents of survey A I found very significant (p<.001) differences for the emotional responses of the respondents who liked or disliked Mother Theresa. Those who dislike her do not report to feel distressed to hear this (mean of 4.38), while those who like her are distressed (mean of 2.88 see table 7.11).

Table 7. 11. Responses to bad gossip about Mother Theresa and Bill Clinton

The results for sharing this bad gossip with others are for both respondents of survey A and respondents of survey B not significantly different when comparing those who like or dislike either Mother Theresa or Bill Clinton. All respondents report not to share this bad gossip with others.

The last bad gossip story is about Elvis Presley (survey A) and Britney Spears (survey B) being disrespectful towards their fans (story 11 in table 7.1). The control group respondents gave this gossip an average rating of 2.60 on a 1-7 Likert scale. Looking at the results for the emotional responses of the respondents on this bad gossip, again all report distressed feelings. Respondents of survey A who dislike Elvis Presley are significantly (p<.01) less distressed than the respondents of survey A who like Elvis Presley. Respondents of survey B who dislike Britney Spears are also, and even very significantly (p<.001) less distressed than the respondents of survey B who do like Britney Spears. Both results confirm my predictions 1a and 1c.

Table 7. 12. Responses to bad gossip about Elvis Presley and Britney Spears (11)

For the tendency to share this bad gossip with others I found very significant differences (p<.001) for both cases. For both Elvis Presley and Britney Spears as gossipees I found that those respondents who dislike the gossipee have a lower tendency to share this bad gossip with others. This contradicts my predictions 2a and 2c and even shows an opposite trend of what I expected.

For all 18 cases I found 7 cases where significant results showed proof, and one significant but low N case that indicates proof for predictions 1a and 1c. Respondents who disliked the gossipees of a bad gossip reported to feel less distressed than respondents who liked the gossipees. Out of these 18 cases 6 have clearly disproved and 1 case indicated disproof of predictions 2a and 2c. Respondents who dislike a gossipee of a bad gossip do not tend to share this gossip more easily with others than respondents who do like the gossipee.

In conclusion of the negative gossip stories I here discussed, I notice that respondents feel distressed to hear bad gossip about celebrities, whether they like them or not. There is some tendency that those respondents who dislike the gossipee feel a little les distressed than those respondents who do like the gossipee, still the differences are small and often non significant. The results for the tendency of the gossipees to share these negative gossip stories with others do not at all confirm my predictions. I had expected that those who dislike the gossipees of such negative gossip to be more willing to share this with others, still my results indicate an opposite trend. For almost all cases I noticed that respondents who like the celebrity gossipee are more willing to share the bad news with others than respondents who dislike the celebrity gossipee. Still, almost all results do not reach an average mean higher than 4.00, which still indicates more a tendency not to share the gossip with others than a tendency to share the bad gossip with others.

4.3.2 Good gossips about liked and disliked celebrities

I now turn to the positive gossip stories that increase the status of the celebrity gossipees. I will start with the most positive ranked stories by the control group. The story that was rated most positive of all 20 was the gossip story about a gossipee who saved the life of a young boy drowning in a lake (gossip 8 in table 7.1). This story got an average rating of 6.28 on a 1-7 Likert scale by the control group respondents. Gossipees of this gossip were Bill Clinton (survey A) and Marilyn Monroe (survey B). When looking at the results for the emotional responses of the respondents (see table 7.13) I notice that, in line with my predictions, those who like Clinton or Monroe are highly significant (p<.001) more happy to hear this good gossip about them than those respondents who do not like Clinton or Monroe.

The results for the tendency to share good gossip about Clinton or Monroe are also in line with my predictions. Those respondents who like Clinton or Monroe are significantly (p<.001 for Clinton as gossipee and p<.01 for Monroe as gossipee) more willing to share this good gossip story about them with others than those respondents who do not like Clinton or Monroe.

Table 7. 13. Responses to good gossip about Bill Clinton and Marilyn Monroe

The second most positive gossip story according to the control group respondents is the story about the gossipee who saved the life of a young girl, beaten up by some guys at school (story 10 in list 7.1). Average rating by the control group for this gossip was 6.25 on a 1-7 Likert scale. Gossipees were Fidel Castro (survey A) and Saddam Hussein (survey B). Similar as with the respondents of survey A, who had to rate how much they liked/disliked Saddam Hussein, the respondents of survey B, who reported to like him are very few in number. Only two respondents to be exact, indicated a strong liking for Hussein. Because this group is so small, it is hard again to draw conclusions from these two respondents, and the results for the “Like Hussein” group are merely and indication. Looking at the results for the emotional responses to hearing this good gossip, all respondents who liked or disliked Castro or Hussein reported to feel rather positive. In line with my predictions, the respondents who like Castro or Hussein indicate slightly more positive feelings, but for both the differences are non significant and, again, the results for Hussein are merely an indication.

Table 7. 14. Responses to good gossip about Fidel Castro and Saddam Hussein

When looking at the results for sharing this good gossip with others, I again notice that those respondents who like Castro or Hussein have a slightly stronger tendency to share this information with others. For gossipee Castro this difference is significant (p<.05), for gossipee Hussein the differences in not significant, and the number of respondents who like Hussein is too small (N=2) to draw conclusions.

The third most positive gossip story according to the respondents from the control groups concerned the story about the gossipee who supported victims from a fire by helping a hand rebuilding their houses. (story 17 in table 7.1). The average rating of this good gossip by the control group was 6.00 on a 1-7 Likert scale. Gossipees of this story were Arnold Schwarzenegger (survey A) and President Bush Jr. (survey B). Again the results (see table 7.15) show that all respondents are rather happy to hear this gossip story about either Schwarzenegger or Bush Jr. as gossipee. For the gossip story with Schwarzenegger as gossipee those respondents who like Schwarzenegger even significantly (p<.01) show stronger positive reactions than those respondents who dislike Schwarzenegger. I did not find significant differences for the emotional reaction of those respondents who liked or disliked President Bush Jr. However, here again the number of respondents who like Bush Jr. is extremely low (N=5). Because of this low number of respondents the results for the story with Bush as gossips should be interpreted as an indication rather than a result. Still the results for Schwarzenegger as gossipee confirm my hypotheses 1b and 1d.

Table 7.15. Responses to good gossip about Arnold Schwarzenegger and President Bush Jr.

Looking at the results for the reported tendency to share this good gossip story with others, I notice a stronger tendency to share this gossip for respondents who like Schwarzenegger or Bush Jr. than for respondents who dislike these celebrities. Again the differences for Schwarzenegger as gossipee are highly significant (p<.001), while the results for Bush Jr. as gossipee are not significant and should be interpreted with much care again, due to the low number of respondents. The results with Schwarzenegger as gossipee again confirm my hypotheses 2b and 2d.

Turning to the fourth most positive gossip story according to the control group respondents, I now discuss the results for the story about the gossipee who cares for homeless people, gives them money and even invites homeless people over at home (story 12 in table 7.1). This good gossip got an average rating of 5.90 on a 1-7 Likert scale by the control respondents. Gossipees for this good gossip story were Prince Charles (survey A) and Mother Theresa (survey B). Al respondents indicated rather positive feelings when confronted with this story. For both Prince Charles and Mother Theresa as gossipee I found significant (p<.05 for Charles as gossipee and p<.001 for Mother Theresa as gossipee) differences between respondents who liked or disliked Charles or Mother Theresa (see table 7.16). In line with what I predicted in hypotheses 1b and 1d, those respondents who like the celebrity gossipee of a good gossip story feel more positive when confronted with this good gossip story than those respondents who do not like the gossipee.

Table 7.16. Responses to good gossip about Prince Charles and Mother Theresa

For the results on the respondents’ tendency to share this good gossip story with others I again find significant results supporting my hypotheses 2b and 2d. Both the respondents who like Prince Charles and the respondents of survey B who like Mother Theresa are more willing to share this gossip with others than the respondents of survey A who dislike Prince Charles and the respondents of survey B who dislike Mother Theresa (resp. p<.01 and p<.001).

Discussing the results of the fifth most positive rated gossip story I now turn to the gossip story about the gossipee who visited children and their parents in children’s hospitals (story 9 in table 7.1). The average rating of the control group for this good gossip is 5.88 on a 1-7 Likert scale. Gossipees of this charity gossip story were Britney Spears (survey A) and Elvis Presley (survey B). Results (see table 7.17) show that for the emotional responses again respondents of survey A who like Britney Spears and respondents of survey B who like Elvis Presley show highly significant (p<.001) positive reactions to receiving this gossip story than respondents of survey A who disliked Britney Spears and respondents of survey B who disliked Elvis Presley. Again this confirms my hypotheses 1b and 1d.

Table 7. 17. Responses to good gossip about Britney Spears and Elvis Presley

Also for the results on the question on how likely the respondents would share this gossip story with others, I find support for my predictions. Both those respondents from survey A who liked Britney Spears and those respondents of survey B who liked Elvis Presley are highly significant (both p<.001) more willing to share this gossip story with others than those respondents of survey A who disliked Britney Spears and those respondents of survey B who disliked Elvis Presley.

I now discuss the results for the gossip story about the gossipees who are very forgiving and do not get upset easily (story 14 in table 7.1). This story got an average rating of 5.6 by the control groups, which is still a strong positive quotation. Gossipees of this gossip story were Grace Kelly (survey A) and Princess Diana (survey B). Looking at the results (see table 7.18) I notice that the respondents of survey A who like Grace Kelly and the respondents of survey B who like Princess Diana report significant (p<.01 for Grace Kelly as gossipee and p<.001 for Princess Diana as gossipee) stronger positive feelings when hearing this gossip information than the respondents of survey A who dislike Grace Kelly and the respondents of survey B who dislike Princess Diana. This again confirms my predictions 1b and 1d.

Table 7.18. Responses to good gossip about Grace Kelly and Princess Diana

Also for the results on the tendency to share this good gossip with others I find support for my hypotheses 2b and 2d. Respondents of survey A who like Grace Kelly and respondents of survey B who like Princess Diana are have a highly significant (both p<.001) stronger tendency to share this good gossip with others than the respondents of survey A who dislike Grace Kelly and the respondents of survey B who dislike Princess Diana (see table 7.18).

The third last good gossip story concerns the information that the gossipee donates a large amount of money to the WWF to save animals (story 15 in table 7.1). Gossipees were Ronald Reagan (survey A) and Brigitte Bardot (survey B). Again all respondents report positive feelings in reaction to receiving this information (see table 7.19). For Ronald Reagan as gossipee I did not find significant results between the respondents of survey A who liked him and those respondents who disliked him. For Brigitte Bardot as gossipee I did find significant differences between the respondents of survey B who reported to like her and the respondents of survey B who reported to dislike her. The first score strongly significant (p<.001) higher than the latter (5.86 vs. 4.72; see table 7.19). Overall the results for this story partly confirm my predictions 1b and 1d.

Table 7.19. Responses to good gossip about Ronald Reagan and Brigitte Bardot

For the results on the tendency to share this good gossip story with others I find support for my predictions 2b and 2d for both Ronald Reagan and Brigitte Bardot as gossipee. Respondents of survey A who liked Ronald Reagand and respondents of survey B who liked Brigitte Bardot showed a highly significant (both p<.001) stronger tendency to share this good gossip story with others than respondents of survey A who disliked Ronald Reagan and respondents of survey B who disliked Brigitte Bardot (see table 7.19).

The second but last good gossip story is about Adolph Hitler (survey) and J.F Kennedy (survey B) who are being good hosts, organizing nice dinner parties and spoiling their guests. This kind of information was rated with a mean score of 5.53 by the N=40 control group respondents, and thus classified as good gossip. The results for the emotional responses of the respondents again confirm my predictions 1b and 1d (see table 7.20). However, I have to remark again that the number of people liking Hitler is extremely low (N=2), and that the results with Hitler as gossipee can only be interpreted as an indication of support for my predictions. The results for J.F Kennedy as gossipee (survey B) can support my predictions. Respondents of survey B who like J.F. Kennedy show highly significant (p<.001) more positive reactions to hearing this good gossip than the respondents of survey B who dislike J.F. Kennedy (see table 7.20).

Table 7.20. Responses to good gossip about Adolph Hitler and J.F. Kennedy

The results for the respondents of survey B on the question whether they will share this good gossip with others or not also confirm my predictions. The respondents of survey B who like J.F. Kennedy are highly significant (p<.001) more tended to share this good gossip story with others than the respondents of survey B who do not like J.F. Kennedy. For the respondents of survey A, with Hitler as gossipee of this same story I found similar highly significant (p<.001) results, but again, since the number of respondents liking Hitler is so extremely low (N=2) I only regard this result as an indication of proof for my predictions.

To end this overview of the results for the good gossip stories I conclude with the results for Michael Jackson (survey A) and Madonna (survey B) as gossipees of the information that they are nice pop stars, responding fan mail and trying to make their fans happy (story 1 in table 7.1). Again results for the emotional responses support my hypotheses 1b and 1d (see table 7.21). Respondents of survey A who liked Michael Jackson and respondents of survey B who liked Madonna show highly significant (p<.001) stronger positive reactions to receiving this information than respondents of survey A who disliked Michael Jackson and respondents of survey B who disliked Madonna (see table 7.21).

Table 7. 21. Responses to good gossip about Michael Jackson and Madonna

Also the last results for the tendency of the respondents to share this good gossip with others confirm my predictions. The respondents of survey A who dislike Michael Jackson and the respondents of survey B who dislike Madonna are not very willing to share this good gossip about these celebrities with others (mean1= 1.69 and mean2= 1.52). This is for both groups highly significant (both p<.001) lower than the tendency to share this good gossip of the respondents of survey A who like Michael Jackson (mean=4.25) and the respondents of survey B who like Madonna (mean=3.5). This last mean score does however indicate rather a tendency not to share than a tendency to share, because of the low mean score of 3.50.

From all 18 good gossip cases 14 significantly proof predictions 1b and 1d, and another 14 cases significantly proof predictions 2b and 2d. Which overall is a support for these predictions. In conclusion of the results for the positive ranked gossip stories I can conclude that my hypotheses 1b, 1d, 2b and 2d can be confirmed. As I expected respondents who like a gossipee of a good gossip story feel more positive when finding out about this than respondents who do not like the gossipee of a good gossip story. In line with this, respondents who like the gossipee of a good gossip story report to be more willing to share this information with others than respondents who do not like the gossipee of a good gossip story.

5 Conclusion

In this paper I tested whether respondents are more likely to share good gossip about celebrities they like and bad gossip about celebrities they dislike. My predictions are funded on McAndrew and Milenkovic’s (2003) results that for real life friends and foes people seem to like good gossip about their friends and like bad gossip about their foes. We share more good than bad gossip about our real life friends with others and we share more bad than good gossip about our foes with others. I term this kind of gossip use Ally Maintenance Reputation Gossip. Increasing the status of allies and decreasing the status of non-allies is an adaptive strategy since the relative status of the individuals who uses Ally Maintenance RG increases.

I combined these predictions with the hypothesis that says that some regard celebrities as one-way members of their social network. For those who regard some celebrities as one-way friends and one-way foes, I therefore expect the same functioning of Ally Maintenance RG to occur.

Presenting 942 respondents with 18 celebrities whom are well known enough to be regarded as one-way members by many, I asked my respondents to indicate how much they liked or disliked these people. I only took into account extreme likings and extreme disliking reactions of the respondents. I expect those respondents who extremely like a celebrity to regard this celebrity as a one-way friend. Likewise I expect those respondents who extremely dislike a celebrity to regard this celebrity as a one-way foe. Next, I presented part of the respondents some good gossip stories and bad gossip stories about some celebrities. For each celebrity I presented a bad gossip story for in survey A, I presented a good gossip story in survey B, and for the celebrities I used as gossipees for bad gossip stories in survey A, I presented good gossip stories in survey B.

Respondents were asked to rate how they felt when hearing the good or bad gossip story about each celebrity on a 1-7 Likert scale with “1” feeling distressed and “7” feeling happy to hear this. Additionally I asked them to indicate to what extent they were likely to pass on this information to others. I again used a 7-point Likert scale with “1” meaning “not at all” and “7” very much.

Linking the like/dislike attitude to the emotional responses and tendency to pass on the gossip information I conclude that for good gossip stories respondents who like the gossipee indeed feel happy to hear the information and are likely to pass this on to others. Also, respondents who did not like the celebrity gossipees of good gossip stories fell less happy to hear this information and were less likely to pass on the good gossip stories. This is in line with McAndrew and Milenkovic’s (2002) results and my predictions.

However for the bad gossip stories I did not find similar results as McAndrew and Milenkovic (2002) and I could not confirm my predictions. The emotional responses of the respondents somewhat indicated that those who dislike the gossipees of bad gossip stories do feel less distressed to hear bad gossip stories about these gossipees that those respondents who report to like the gossipees. However the results for sharing bad gossip stories with others not at all confirm my predictions. The opposite from what I predicted follows from my results. Respondents who like the gossipees of bad gossip stories are more likely to pass on this information to others than respondents who do not like the gossipees of bad gossip stories.

Overall it follows from this study that respondents who like the celebrity of a gossip story are more likely to pass on this gossip information to others than those respondents who dislike the celebrity gossipees of good and bad gossip stories.

6 Discussion

Generally good gossip stories are more likely to be passed on to others than bad gossip stories. McAndrew and Milenkovic (2002) suggested that we should be more eager to spread bad gossip about higher status others to relatively increase our own status does not get support from this study. I have argued that I do not go along with their prediction about this. I rather think we will share good gossip about allied high status others and bad gossip about non-allied high status others. My results indeed show that good gossip about high status celebrities is more likely to be shared with others. We can bask in the glory of high status allies if we praise them, which might be the case for those who regard celebrities as one-way members of their social networks.

I also see a link to the results of my content analysis I presented in paper 4. I there concluded that more positive than negative gossip is written about celebrity gossipees. I then explained that this makes sense from the viewpoint of the Learning Hypothesis. If we are interested in celebrity gossip because we can learn from this information which strategies to mimic in order to become successful, it makes sense that we are more interested in the successes of these people than in their failures. The trend that follows form the results stemming from this study also indicates that positive gossip about celebrities has higher chances to be passed on to others.

I think future research should concentrate on this difference. It would also be nice to investigate whether these successes of celebrities are passed on to just anyone, or only to friends and allies.