The Aggressiveness of Playful Arguments
- First Online:
- Cite this article as:
- Hample, D., Han, B. & Payne, D. Argumentation (2010) 24: 405. doi:10.1007/s10503-009-9173-8
- 163 Views
Some people report that they argue for play. We question whether and how often such arguments are mutually entertaining for both participants. Play is a frame for arguing, and the framing may not always be successful in laminating the eristic nature of interpersonal argumentation. Previous research and theory suggest that playfulness may be associated with aggression. Respondents (N = 199) supplied self-report data on their arguing behaviors and orientations. We found support for the hypothesis that self-reported playfulness and aggression are directly associated. We found less evidence for our hypothesized inverse association between self-reported playfulness and indices of cooperation and avoidance. Self-reports of playfulness are not significantly associated with expert coders’ ratings of either playfulness or aggressiveness. The claim that an argument is playful should be met with skepticism, although playful arguments are possible.
KeywordsArguingArgument framesPlayAggressivenessCooperationArgumentativenessVerbal aggressiveness
Although many argument theorists see the goal of arguing as persuasion or consensus, arguments can in fact be undertaken for a variety of personal reasons. Someone may argue to display a positive identity (e.g., a student writing an answer on an essay exam), to project dominance over another person (e.g., a supervisor giving empty reasons to do what s/he says), to generate an example (e.g., an instructor illustrating an argument for analysis), or to describe a relationship (e.g., two actors performing a script to show why the fictional couple doesn’t get along). This is not an exhaustive list of the reasons an argument might appear in interaction, and this paper concerns another such reason for arguing: to play.
While not everyone immediately understands that arguing can be undertaken for fun, some people do engage in the activity for entertainment. Interscholastic debaters are one example. More pertinently to this paper, some relationships contain continuing arguments about sports or politics or some other topic, and the partners return to the issue frequently to pass the time without any expectation or wish that the argument might ever really conclude. Some people, as we will see, report that they start arguments for fun: they make outlandish remarks or poke at people to see what they will say or to begin a pleasurable exchange. When both partners share the same frame for the argument—that it is for entertainment—the exchange has the potential to be mutually rewarding. When one of the participants does not consider the argument to be playful, the interaction is far less likely to be mutually enjoyable.
The purpose of this investigation is to analyze and question the playfulness of arguments that are so labeled by at least one of the participants. We propose that play and aggression are closely linked, both theoretically and empirically. While pure cases of playful argument are possible, we will test the general hypothesis that they are more rare than the participants suggest, and that we must be wary of the claim that any particular argument is playful. We begin with a theoretical basis for our study, and then proceed to review the empirical findings that are on point.
1 Goffman on Frames and Hample on Argument Frames
While an argument is a real thing, an objective event in the social world, it is also a subjectively understood one. Goffman’s (1974) idea of frames captures this duality, and Hample (2003, 2008) applies the concept of frames to interpersonal arguing.
What I encountered at the zoo was a phenomenon well known to everybody: I saw two young monkeys playing, i.e., engaged in an interactive sequence of which the unit actions or signals were similar to but not the same as combat. It was evident, even to the human observer, that the sequence as a whole was not combat, and evident to the human observer that to the participant monkeys this was “not combat.” (p. 179, italics original)
While the monkeys performed combat actions, they privately framed those behaviors not as fighting but as play, and this framing was observable to Bateson.
Beginning at this point, Goffman (1974) developed first a description of play frames and then a theory of framing in general. His overall understanding of a frame is that it is both a personal understanding of some social event and the observable episode itself. These need not be the same. For instance, a confidence swindler operates by projecting a false frame for his/her victim so that the victim’s private understanding does not correspond to what is actually going on. The victim is said to be “contained” in the swindler’s frame in such a case. Jokes, especially practical jokes, have a similar character. Containment can also occur when an episode is misunderstood by a participant and he or she is thereby unable to participate properly in the interaction.
The main points of Goffman’s theory are these. First there is a “natural strip” of behavior. This is its essence, its base appearance and character. A natural strip can be transformed. For instance, passing a valued object can be transformed into a gift, and a gift can be retransformed into a bribe, an unwelcome advance, or a promise of marital fidelity. Transformations are done by reference to keys. By “key” Goffman has in mind a musical metaphor, as when a melody is put into another key to make it livelier or moodier. Another way to understand key is by reference to cryptology: a message can be keyed (i.e., put into code) so as to appear very differently. Transformations are controlled by their keys, and they result in laminations. A lamination is a new appearance for the natural strip and it is the lamination that defines the event for participants and observers. Laminations cover behaviors in the same way that a good veneer can cover inexpensive wood. Laminations are supposed to be opaque so that the event genuinely becomes only what the lamination appears to be, and nothing of a more basic nature is evident or in social use.
Many layers of lamination are possible for a single natural strip of behavior. For example, on the old Honeymooners television show the lead male actor would often pantomime a huge uppercut blow aimed at his wife and shout “To the moon, Alice, to the moon,” conveying that he would hit her so hard that she would fly to outer space. The natural strip of behavior here is a combat move, a punch. It is re-keyed into an exaggerated blow, obviously useless in a real fight. Then it is transformed to be in a theatrical production, so that the audience understands that no real person is in danger. It is transformed once more because the context of the show makes it clear that he would never actually strike his wife. And it is transformed one last time to make it funny, since this was considered hilarious in the 1950s. Only that last lamination was really observed and reacted to by the audience. So we would say that they framed such an episode as comedic.
Goffman summarizes his understanding of frame in this way: “I assume that definitions of a situation are built up in accordance with principles of organization which govern events—and our subjective involvement in them; frame is the word I use to refer to such of these basic elements as I am able to identify” (Goffman 1974, pp. 10–11). Thus when people understand a social episode they attend to the rules that define and regulate particular sorts of interaction and also relate what they observe to their own place in the episode. They define a situation, they frame an event, they directly experience a social moment. This interpretive process involves a subjective perception of what is also acknowledged to be a public social phenomenon. Frames are simultaneously internal and external, private and public. When interaction goes well, the internal and external frames match. When something is off, there may be a mismatch, which we might eventually recognize as a swindle, a misunderstanding, a put-on, a lie, or a false front.
Working with a less developed conception of frames, Hample (2003) addressed the question, “What do ordinary arguers think they’re doing when they argue?” His answer developed a variety of frames or perspectives that people seem to apply when they argue. His argument frames are in three categories.
The first set of frames is self-oriented and relegates the other arguer to the background, as merely being a resource or foil needed to achieve one’s individual goals. Within this first set are four frames: utility, identity, dominance, and play. A person may argue for a utilitarian reason, for instance to obtain a favor. An argument might be initiated to display a particular aspect of one’s identity, as when a person argues elaborately for a political position that his/her listeners are already known to endorse. Dominance can be the motive for arguing when one person intends to oppress another by force of status, knowledge, or arguing skill. And finally someone might argue for play, just to have the entertainment of a spirited exchange.
The second set of frames (which Goffman would probably call keys rather than frames) takes the other arguer’s goals into account, promoting the cointeractant from the background to the status of a full participant. The transition from the first set of frames to the second is captured by the idea of blurting, or speaking without editorial work. Those who blurt do not really arrive at this second general orientation. They take pride in saying whatever is on their mind. But those who do not blurt have taken the other person into account and have tried to adapt to his or her needs and wishes. The next issue in this second category of frames is whether adaptation to the other is competitive or cooperative. Either possibility can be genuinely adaptive, but cooperation is more advanced. Seeing an argument as cooperative means that the interlocutor’s needs are put on a par with one’s own, rather than being subordinated or exploited as happens in competition. The final matter here is called civility, the perception that an argument is polite and respectful of the other. Civility is also a transitional measure, this time referring to the possibility that a person achieves the third and most advanced set of frames.
This third “set” is actually comprised of one group of understandings, collected under the label professional contrast. Argument professionals (scholars are the implicit reference group) understand that arguing is an alternative to violence, that it is productive, that it can improve relationships and understanding, and so forth. Many ordinary actors believe the reverse and these arguers display considerable contrast to the views of professionals. Respondents who have high scores on this measure (i.e., they agree with argument professionals) are regarded as having the most advanced understandings of arguing.
Moving through the three general categories of frames, we see a conceptual progression from self orientation to genuine engagement with another person to a reflective understanding of argumentation in the abstract. Hample speculates that the three categories may also be in developmental order, but no data have been collected to test that supposition.
Obviously the play frame is of special interest here. Several empirical studies have related this frame to the others, as well as to some other conceptually related constructs. Self-report scales have been developed to display the degree to which respondents see the possibility that arguments can be playful (or utilitarian, or cooperative, and so forth). Hample (2005) found that sensitivity to the possibility that arguments can be playful was positively correlated with recognizing identity uses (r = .47) and dominance impulses (r = .44), but was more associated with competition than cooperation expectations (r = −.45). The play frame also correlated with masculinity (r = .48), femininity (r = −.40), argumentativeness (r = .50), and verbal aggressiveness (r = .31). These are initial grounds for suspicion that argumentative play may not be pure: if it were, why would it be connected so strongly to competition, verbal aggressiveness, masculinity, or arguing to display dominance? Hample et al. (2009) replicated several of these results. Play correlated significantly with identity (r = .56), dominance (r = .41), competitiveness (r = −.14), masculinity (r = .25), femininity (r = −.20), verbal aggressiveness (r = .35), argument-approach (r = .52), argument-avoid (r = −.44), and psychological reactance (r = .52). Again we see the recognition that arguments can be playful being associated with a group of indices that suggest aggressive interpersonal combat. Another sort of evidence comes from Hample et al. (2006), who asked respondents to rate sample arguments as to their playfulness and appropriateness. These correlations were negative (r = −.16) for personal topics, and nonsignificant for public topics, which would seem to offer more opportunity for low-stakes play than for personal matters.
So empirical evidence to this point associates sensitivity to play with a variety of measures indicating aggression and uncivil interpersonal relations. The only apparently countervailing evidence is that two studies indicated a positive relationship between liking for conflict and the play frame (r = .45; Hample 2005) and between liking for conflict and ratings of arguments as playful (r = .15 for personal topics and r = .14 for public topics; Hample et al. 2006). One would suppose that brutish arguments would not be liked. Perhaps the aggression is precisely what produces positive valence for interpersonal arguing in some people.
But laminations are supposed to be opaque. Properly keyed, when an argument is offered in a playful way, that should be its only apparent nature. We ought not be seeing associations with perceptions and expectations that, on their face, are not entertaining, playful, or fun. Perhaps the play lamination does not work properly.
This possibility requires that we think seriously about a basic issue: In the case of arguing, what is the natural strip of behavior? In the analysis of many social activities this will not be an important sort of question. Even if we mistake, say, the third lamination for the natural strip, as long as all the laminations are opaque we will still characterize the episode correctly by reference to its top lamination (perhaps the fourth or fifth). But if the lamination is not entirely opaque and part of what lies underneath asserts itself, we need to know what the veneer is supposed to be covering.
We are now inclined to believe that the natural strip of behavior—the strip for which any other sort of argument is a rekeying—is verbal combat. For many purposes seeing a demand or the exchange of reasons as the natural strip might suffice (e.g., Hample et al. 2009), but here we wish to make more basic points. An articulate version of a fight or marauding raid, an argument is a verbal way of beating someone up and taking something away.
David: Hey. Let me have that pig.
Chris: That’s my pig!
David: No. Let me see that pig okay? You no//
Chris: Give me my pig! [Screams]
David: Let me have it okay?
Chris: No. I want this. I want that pig.
David: No. That’s my pig.
Obviously we see nothing here of civil adaptation, no mutual journey to enlightenment, no effort to alter the other except by verbal force. This is the natural strip, an eristic struggle in which reasons and reasonableness are alien.
Arguments don’t simply appear. They are capacities and resources that are used, and they are done in service of some personal goal. The advanced goals are things like knowledge seeking, mutual accommodation, and mutual entertainment. But the base goals are obtaining some resource or establishing a continuing right to overwhelm the other person. Advanced goals are pursuable only when lower level strips of behavior have been rekeyed, set aside, and successfully laminated. In the case of playful arguing, we propose that the lamination is often unsuccessful and that cruder and more aggressive features of arguing are not properly veneered.
2 Specific Variables
In order to test this general hypothesis and to promote more detailed analysis of playful arguing, we undertook a study with a substantial number of variables. An important improvement on the earlier studies of playful argument is that we have collected examples of arguments that participants identified as playful. We have also supplemented the frames measure (intended to show sensitivity to the possibility that arguments can be used for entertainment) with another scale that asks respondents to report the degree to which they actually argue for fun. We have done a similar thing for cooperation and competition, so that we have measures of whether a respondent thinks that arguing can be cooperative as well as whether the person reports that s/he argues cooperatively himself/herself. In other respects, we have made use of variables that have been previously studied in regard to the play frames measure as well as some others, to replicate and extend the empirical record, and to trace out relationships with the new play measures.
RH1: The three measures of playful arguing (the frames measure, the self-report of playful arguing, and coders’ play ratings of self-identified playful arguments) are positively correlated.
RH2: The three measures of play are positively associated with variables associated with aggressiveness: the antisocial subscale of the verbal aggressiveness instrument, masculinity, reactance, positive valence for conflict, the dominance frame for arguing, the approach subscale of the argumentativeness instrument, people’s self-reports of how competitive they are, and our ratings of the aggressiveness of the open-ended examples of playful arguing.
RH3: The three measures of play are negatively associated with variables associated with avoidance and cooperation: communication apprehension, femininity, the prosocial subscale of verbal aggressiveness, the cooperative frame for arguing, self-reports of cooperative arguing behavior, the civility frame for arguing, and the argument-avoid subscale of argumentativeness.
The data for this study were collected in large part for pedagogical purposes. Students enrolled in an upper division theory construction course taught by the authors provided the data in order to write reports on the topic of playful arguing and to develop their own theories. Surveys were completed online. Students were encouraged to suggest additional measures and they did. Some relevant measures only appeared in the second semester (the course was taught twice). Many student-suggested variables are excluded from this report. A full list is available from the first author. Students also gave consent for their data to be used for research purposes, and this study was approved by our Institutional Research Board.
A total of 199 respondents provided data. With a few exceptions, they were communication majors at our institution. They were spread over the two semesters, and data on some variables are missing because not all measures were used in both terms. Only in the second semester were students asked to supply an example of a playful argument, and sample size is 99 for measures based on those reports. 28% of respondents were male and 69% were women, with 5 participants not responding to this question. No freshmen were in the sample; 13% were sophomores, 50% juniors, and 34% seniors. 45% of the sample self-identified as being of European ethnicity, 7% African, 7% South or Central American, 5% East or Central Asian, 28% other (perhaps including ethnic combinations), and the remainder were scattered among various ethnic origins or preferring not to answer. Respondents averaged 20.7 years old (SD = 1.6).
Three measures focused on the idea of play in arguments. The first is the play scale from the frames instrument (Hample et al. 2009). This measures awareness that arguing can be done for fun, and is a four item Likert measure with Cronbach’s alpha = .83. The second is the self-report that the respondent actually does argue for entertainment. We created sets of five Likert items to measure the frequency with which one argues playfully in general, with superiors, with friends, and with strangers, but found all these so intercorrelated that we combined them into one measure (“play self-report”), with Cronbach’s alpha = .80. Third, two of us rated the examples of playful arguments on four Likert items reflecting the apparent entertainment value of the episode, with an association between the coders of r = .75. For the first coder, the Cronbach’s alpha for the four items (playful, enjoyable, entertaining, and cooperative) was .98. The second coder’s Cronbach’s alpha was .99. Combined means for the two raters were used as the playfulness rating.
Several measures are collected here under the general heading of aggression, which we understand to mean willingness to hurt the other in some way, a commitment to getting one’s own way even at the expense of the other, or an impulse to approach conflictual interaction. The open-ended reports of playful arguing were coded by two of us for their aggressive character. For the first coder, Cronbach’s alpha for the four items (one-sided, hurtful, aggressive, and competitive) was .96. The second coder’s Cronbach’s alpha was .93. The two coders’ ratings correlated at r = .77. Combined means for the two raters were used as the aggressiveness rating. The verbal aggressiveness scale (Infante and Wigley 1986) has two subscales (Levine et al. 2004) that measure self-reports of antisocial and prosocial behaviors. The antisocial scale is the aggressiveness measure, and it had a Cronbach’s alpha of .84. Masculinity (Bem 1974) represents a forceful identity. Using Likert responses, we obtained a Cronbach’s alpha of .84. Psychological reactance (Dowd et al. 1991) is a 28 item self report scale that measures one’s inclination to respond aggressively when an interlocutor tries to constrain one’s freedom to act or think, and here it had a Cronbach’s alpha of .80. Positive valence for conflict is one of the subscales of the taking conflict personally battery (Hample and Dallinger 1995) and it measures the degree to which one enjoys and approaches interpersonal conflict. Here the seven item measure had a Cronbach’s alpha of .81. The dominance frame for arguing (Hample et al. 2009) measures the respondent’s sensitivity to the possibility that a person may argue in order to exert dominance over the interlocutor. Here the six item scale had a Cronbach’s alpha of .76. The next aggression measure is the argument-approach subscale of the argumentativeness scale (Infante and Rancer 1982), and its Cronbach’s alpha was .85. Respondents self-reported their general levels of competitiveness during arguments on seven Likert items, and this scale produced a Cronbach’s alpha of .78.
3.2.3 Avoidance and Cooperation
Here we have collected measures that indicate either that one wishes not to participate in arguments or that, once engaged, the main key is a cooperative, unaggressive one. The first measure in this category is the PRCA-24 (McCroskey 1978), which measures communication apprehension. Its Cronbach’s alpha was .94. The femininity subscale of the BSRI (Bem 1974) reflects a mild, nuturant identity and was measured with Likert items; the Cronbach’s alpha was .81. The prosocial subscale of the verbal aggressiveness instrument (Infante and Wigley 1986; Levine et al. 2004) reflects people’s propensity to do collaborative face-respecting actions during arguments. Here the measure yielded a Cronbach’s alpha of .78. The cooperation frame (Hample et al. 2009), scored so that high numbers show appreciation for cooperative possibilities in arguing rather than competitive ones, gave a Cronbach’s alpha of .70 when the final item was omitted. Respondents self-reported their general levels of cooperativeness during arguments on five Likert items, and this measure generated a Cronbach’s alpha of .63. The civility frame for arguing (Hample et al. 2009) measures one’s perceptions that arguing is a polite, productive activity, and the Cronbach’s alpha was .76. The argument-avoid subscale of the argumentativeness instrument (Infante and Rancer 1982) produced a Cronbach’s alpha of .85.
3.2.4 Descriptive Statistics
Sample sizes, means, and standard deviations
Correlations among the play measures
On contemplating these results a complication occurred to us. Perhaps people who don’t argue playfully were at a loss when asked to supply an example of a playful argument, and that is why our ratings did not correspond with the self-reports. So we divided our sample, using a median split (at 2.75) on the self-report of playful arguing and compared the two groups as to our ratings of their episodes. We did not quite find support for our suspicion (t (97) = 1.72, p = .09, MeanHI = 3.40, MeanLO = 3.00; power = .80). Correlations between the playfulness self-report and the play ratings for the two groups remain insignificant: r = −.14 (N = 45, power = .65) for the low self-report group, and r = −.07 (N = 54, power = .72) for the high self-report group.
Our substantive conclusion is while people are consistent in their self-perceptions about playfulness while arguing, they are almost perfectly imperceptive about whether their arguments are actually playful. This result is consistent with this study’s motivating suspicion, that so-called playful arguments may not be playful at all.
Correlations between play and aggression measures
1. Agg Rating
8. Compet S-R
Canonical correlations between playfulness and aggression measures (unrotated)
R2 from other set
The first root mainly reflects an inverse relationship between the playfulness and aggressiveness ratings of the open-ended examples of playful arguments.1 This reflects the coders’ understandings of playfulness and aggressiveness in public social episodes, not necessarily the respondents’ perceptions (as we saw in the test of hypothesis 1). The Appendix has examples of argument reports that we rated as being high or low on playfulness and high or low on aggressiveness. The two self-reports of possible and actual playfulness of arguments have loadings consistent with the playfulness ratings. These results do not support the hypothesis, which projected a direct relationship between playfulness and aggression.
The second root, which of course has had the variation due to the first relationships omitted, is dominated by the two self-report play measures. They are directly associated with all of the aggressiveness variables, particularly the approach subscale of the argumentativeness instrument, valence, the dominance frame, and reactance.
Hypothesis 2 is supported with regard to respondents’ self-reports. People’s experience of playfulness is directly associated with aggressiveness. When coders who have advanced understandings of interpersonal arguing rate episodes, however, play and aggression are negatively related. We see here a contrast between participants’ framing and that of expert observers. Taking the coders’ perceptions as being more advanced, we can also see the opportunity for a mismatch between private understanding and public performance.
Correlations between play and avoidance/cooperation measures
4. Coop frame
5. Coop S-R
Canonical correlations between playfulness and avoidance/cooperation measures
R2 from other set
Our interpretation of the significant root is that it mainly shows that the two self-report measures of playful arguing have an inverse relationship with the argument avoidance subscale of the argumentativeness measure. This is the dominant effect in Table 6, and supports hypothesis 3 because it indicates that self-reported playfulness has a noticeable negative relationship with argument avoidance. However, the direct relationship between the play measures and civility is in the opposite direction of the hypothesis. Overall, the results in Tables 5 and 6 tend to support hypothesis 3 but are not unequivocal. In fact, it might be reasonably concluded that the primary result is null, since so many of the avoidance and cooperation measures were unassociated with the playfulness variables.
We believe that the natural strip of arguing behavior is eristic, that at its core arguing is verbal force aimed at defeat of the other person. We do not mean to imply that this is the normal appearance of interpersonal argument. Part of the socializing process involves teaching children to frame their arguments cooperatively. So we are taught that we can learn things from a good argument, that we can clear the air, that we can persuade and stand up for ourselves without hurt feelings, that we can resolve problems and improve our relationships—if we argue with open minds and open hearts. Many parents have the goal of teaching their children not to initiate the natural strip and to frame arguments in a productive way. Rekeyed arguments are the norm in civilized discourse.
Among the possible laminations of the natural strip is arguing for fun. Playful arguing has the potential to provide mutual entertainment—provided that the argument is really playful and that both participants have the same frame and impulse. The results of this study indicate that these possibilities are far more remote than our respondents realize. Laminations are supposed to be opaque, but we believe that this one fails more often than not.
Our first clue that the play lamination is somewhat transparent was that people’s self-reports of sensitivity to play and participation in playful arguments did not predict our own ratings of the playfulness of their arguments. Respondents self-identified these episodes as playful, but our ratings were no higher for the arguments proposed by people who often play than for those who do not. Surely one would expect that people more sensitive to the possibility of play would have generated more obviously playful arguments, and that people who often argue playfully would have had a good example at hand, but this was not so. In fact, our ratings were almost perfectly uncorrelated with people’s self-perceptions.
Based on previous theory and empirical results we made two hypotheses about the relationship between play and aggression, one more conservative than the other. One prediction—the more straightforward one—was that the play measures would be positively associated with various aggression indices. This hypothesis was clearly supported insofar as our respondents’ self-reported experiences and perceptions are concerned. Entertainment is not normally supposed to be eristic or potentially unpleasant, but our results show that in the case of arguing, it certainly is. Aggressiveness asserts itself forcefully in the experience and awareness of arguing for play. The entertainment character of interpersonal arguing is more comparable to boxing than to passing the time pleasantly or working on a garden together. In fact, we are somewhat disinclined to say that playful arguing is playful at all, since it shows such a combative nature in our analyses.
The other hypothesis was the conservative one. If playful arguing is only aggressive then it should be negatively correlated with avoidance and cooperation measures, and that is what we predicted. If playful arguing has a dual nature—partly aggressive and partly cooperative—then it might possibly show some positive connections to civil impulses. We actually found support for both possibilities. Playfulness was negatively associated with argument avoidance, supporting the hypothesis. However, we also discovered a weaker but direct relationship between playfulness and civility, suggesting that argument play is civil to some degree. The argument avoidance result was by far the more powerful one. Most of the avoidance and cooperation measures had little to do with playfulness. Taken together, these results do not present a clear picture, although they provide several important clues about the relationship between playful and cooperative arguing. These should be followed up in future work.
Interactive play should have an essentially cooperative character. A comparison of our results for the aggression and cooperation hypotheses shows that the aggressive nature of playful arguing overwhelms but does not completely nullify the possibility that arguing can be cooperative fun. Even so, it should be noted that our results mainly connected self-reports of playfulness with self-reports of cooperation and avoidance. The coders’ play ratings were not involved in the significant results, so we may only be seeing a naïve theory of playful arguing rather than a reflection of what actually happens, although Goffman’s frames theory reminds us that subjective understandings are consequential. Our ratings of the playful episodes were essentially unrelated to any of the cooperative or avoidance self-reports. We believe that the cooperative nature of playful arguing is modest and overwhelmed by its other nature, the aggressive one.
We do not doubt that arguing can be done for mutual fun. Our playfulness ratings of respondents’ episodes resulted in only 32% being rated below 3, the theoretical midpoint of our scale. 33% were rated at 4 or higher. But we believe that the playfulness is not pure. It is insistently partnered with aggression. The play lamination is not opaque and it is a relationally dangerous sort of rekeying to undertake.
A perceptive reviewer pointed out that an experimenter bias might have been operating here, since the two coders knew that the project’s aim was to explore what we thought was the problematic relationship between playfulness and aggressiveness. While we cannot dismiss this possibility, the high reliabilities of the ratings imply that both coders would have had to exhibit the experimenter bias to almost exactly the same degree, which is somewhat unlikely. A better but more resource-intensive methodology would have been to have the two ratings done by separate sets of coders, all blind to the study’s objectives. Readers can see examples of how the raters distinguished the reports in the Appendix.