Melfi’s Choice: Morally Conflicted Content Leads to Moral Rumination in Viewers

ABSTRACT

This study investigates if morally conflicted and controversial content, which is often denounced as morally desensitizing, may play an important role in leading to moral rumination in viewers. The results of a quasi-experiment reveal that moral rumination was predicted by transportation into the narrative and was related to increased appreciation for the episode. The results support the notion that media potentially function as a “morality sandbox” in which to play with or test out moral concerns, even regarding controversial and violent content.

Introduction

Since its inception, concerns about television content have been closely linked with perceptions of the decay of public morality and social norms (Dant, 2012; Eden, Grizzard, & Lewis, 2013). At the heart of many of these discussions are concerns about violent narratives normalizing violence as a morally acceptable resolution to conflict. Violent narratives on television and film therefore have been the subject of controversy and social criticism as well as scholarly research for decades (e.g., Ferguson & Savage, 2012; Paik & Comstock, 1994). The majority of research has been devoted to studying the potential negative effects of media violence. This follows social learning theories suggesting that media can negatively influence our behavior via our perceptions of what constitutes acceptable behavior, or via our internalization of social reward structures associated with violent behavior (cf. Bandura, 2001).

Particularly concerning in the study of violent mediated content are story lines featuring rape and sexual assault plots, which have seemingly become more prevalent over time (Nussbaum, 2015). For example, The Sopranos episode “Employee of the Month” that featured a graphic, realistic, and unpunished rape of a main female character was seen as a fan favorite and was counted among the most critically acclaimed episodes of the show (Vanderwerff, 2010).

Part of the critical acclaim may stem from the moral ambiguity of the story line; although the sexual assault is the catalyst for the dramatic conflict, the majority of the outrage is in Dr. Melfi’s decision not to punish the criminal (i.e., by telling her client, Tony Soprano, about the assault). Although Tony would exact justice, he would most likely do so via unlawful and violent means, and Dr. Melfi decides in the episode not to tell him of the assault. Therefore, although she upholds the law and her own professional boundaries between therapist and client, she does not serve the narrative “justice restoration” sequence that audiences have come to expect from television. Series creator David Chase aptly illustrated the moral complexity of the episode and also Dr. Melfi’s atypical response when he stated:

If you’re raised on a steady diet of Hollywood movies and network television, you start to think, “Obviously there’s going to be some moral accounting here.” That’s not the way the world works. (Martin, 2007)

Although it seems like such a scenario should be rewritten with proper justifications in mind, moral complexity in media may not be a negative thing. It may allow room for viewers to interact with and conceptualize the media content in unexpected ways. As Hall’s (1993) model of encoding and decoding describes, viewers are not passive viewers of a static text. Instead, viewers shape the meaning of what they see on the screen based on personal frameworks of knowledge and meaning, thus allowing for multiple interpretations of the same act featured on screen. This process of sensemaking invites the viewer to be an active participant in the text based on their own daily life and experiences (Chisholm, 1991; Livingstone, 1990). Furthermore, we interpret media based on our own moral norms and perceptions of moral virtue (Zillmann, 2006); however, when resolution is lacking in the (complex) narrative, we are (actively) forced to consider what we think is right and even what we would have done (cf. Chisholm, 1991). As such, this type of media content may actually play an important role by presenting viewers with complex moral situations that require complex deliberation or moral rumination (Alasuutari, 1992; Dant, 2012; Krijnen, 2007; Lewis, Tamborini, & Weber, 2014; Van Ommen, Daalmans, & Weijers, 2014). This moral rumination may act, in turn, to promote moral education via a reflection on the viewers’ own morality and moral choices (Alasuutari, 1996; Krijnen, 2011). Therefore, it seems that a perspective on complex controversial content, based on the moral response it evokes from viewers, is warranted. In the current study, we explore how the episode “Employee of the Month” affects viewers’ moral rumination, via viewer traits, narrative responses, and personal morality.

Moral rumination

To understand how media and moral rumination are related, first we must define moral rumination. Turiel (1983) suggests that people act as “moral scientists” who construct and simulate social experience to develop their ideas about right and wrong. Moral reasoning, which is often used interchangeably with terms such as moral rumination or moral deliberation, is the “conscious mental activity that consists of transforming given information about people in order to reach a moral judgment” (Haidt, 2001, p. 818). Haidt also suggests that morality functions as a dual process involving both deliberation, or rational reflective thought, and moral intuitions, or quick emotional responses to moral situations. Borg, Sinnott-Armstrong, Calhoun, and Kiehl (2011) distinguish between what we are terming moral rumination—the weighing of moral considerations and moral verdict—and the commitment to a moral conclusion (p. 1) and illustrate that neurologically and conceptually, the two processes are distinct. In their work, moral deliberation is the detection, filtering, and weighing (consciously or unconsciously) of relevant moral principles, heuristics, or concepts that identify morally relevant features and thereby create a “moral context” that takes place over time. In their view, this contrasts with a moral verdict, in which a moral decision is rendered.

In this work, we are interested in both the rational and the emotional side of moral judgments, and are most concerned with the individual’s involvement and engagement with moral issues presented in context. Given the dual usage of the word “deliberation” in prior work, referring both to the thought process as well as the type of thought process being used, in the current article we use the term moral rumination to describe the individual involvement with a moral issue presented in narrative. This rumination may encompass both rational and conscious (i.e., deliberative) processes, and yet also account for emotional responses to moral violation presented in narrative. Thus, we conceptualize moral rumination as the capacity and process by which a person evaluates several perspectives on a moral issue, through which it becomes clear which moral value is most important in a specific situation and what the preferred moral action is going to be (cf. Dewey, 1891).

Importantly, moral rumination can be linked both to moral imagination and moral simulation. That is, individuals do not need to actually be present themselves in a morally demanding situation to think about what they would do or how they would act in a particular situation. This is in line with conceptualizations of media and narrative, particularly those dealing with complex social situations that serve as a “moral playground” or a “moral sandbox” in which viewers can safely try out and practice different responses to social situations (Grizzard et al., 2015; Vorderer, Klimmt, & Ritterfeld, 2004).

Although discussions of media as a moral sandbox have tended to focus on interactive media such as video games, Mar and Oatley (2008) argued that interacting with noninteractive narrative fiction—be it literary or mass mediated—also can be seen as a moral sandbox, in that it encourages empathetic growth and the transmission of knowledge about the social world through the vicarious experience. Mar and Oatley propose that the ultimate function of fiction is the abstraction and simulation of social experience, which helps foster a greater understanding and empathy of ourselves and others. Fictional narratives portray the past, present, and future actions of characters, and as a result we can see their motives, behavior, and consequences with a clarity that is denied us in our own lived experience or personal daily lives (Winston, 1999).

Television narratives, much like literary narratives, offer the viewer a space to work in a moral laboratory (cf. Ricoeur, 1983). In this laboratory, viewers can explore and reflect on moral issues and deliberate without facing the consequences of their moral decisions (Hakemulder, 2000; Krijnen, 2011). Carroll (2000) also proposed that, through fiction, we exercise and augment our potential for moral reflection and moral understanding. Fictional narratives—encompassed in both film and television formats—can contribute to the teaching of moral rules and values (Bivins, 2007; Good & Dillon, 2002; Wilkins, 1987; Winston, 1999), but also aid us in applying known moral rules and values in “concrete cases, engaging and exercising our emotions and imagination, our powers of perceptual discrimination, moral understanding, and reflection, in ways that sustain and potentially enlarge our capacity for moral judgment” (Carroll, 2000, pp. 368−369). Whereas the power of narratives to influence the moral rumination of viewers has mostly been proposed by literary scholars, communication scientists have also explored this possibility. That is, the belief then is that engaging stories that feature characters with whom audiences can identify may help viewers to adopt attitudes centrally contained in these stories (e.g., Green & Brock, 2000; Strange & Leung, 1999).

Regarding moral norms portrayed in media, Klapper (1960) suggested that television content acts to uphold and maintain societal norms. More specifically, research by Alasuutari (1992) showed how following story lines and character development in Finnish television fiction can be seen as a form of moral education for its viewers. Krijnen (2007) qualitatively and quantitatively explored the ways in which prime time television sparked the moral imagination of viewers. She concluded that television, regardless of the genre, sparked moral rumination and moral insight, and should be seen as a resource in the process of becoming a more morally mature individual.

Recently, Van Ommen et al. (2014) studied the role that a morally ambiguous doctor-protagonist (House, M.D.) could fulfill in a medical ethics program. They concluded that contrary to many of the concerns about Dr. House as a bad example for medical professionals, their results showed “that the use of a visual narrative like House, M.D. can create a starting point to access the ‘immaterial culture’ of medical students and starting physicians relating to their moral evaluations about professional conduct” and can have “a meaningful role in helping to teach ethics to medical students” (p. 70). In a similar vein, Dant (2012) argued that viewers do not have to agree “with the moral premises of the characters in a television show and may actually be alerted to the moral issues implicit precisely because he or she does not agree” (p. 42). Finally, Lewis et al. (2014) demonstrated that presenting viewers with conflicted moral scenarios in narrative promoted deliberative processing and a deeper appreciation for the narrative content.

In other words, previous research seems to suggest that television can stimulate moral rumination on the part of the viewer by presenting the viewer with a variety of views on moral topics that provide insight into the human character as well as showcasing behavior that transgresses widely shared norms and values. That rumination on the side of the viewer may lead to a growth in moral maturity, which can be understood as the capacity to assume “the responsibility for working out their own [moral] solutions and making their own [moral] judgments” (Rosenblatt, 1938, p. 201).

The current study

In the current study, a story line centered on a morally complex issue lacking justification or resolution was chosen from the dramatic TV show The Sopranos. The story line revolved around the character Dr. Melfi, from the episode “Employee of the Month” (S03E04) of The Sopranos. This story line features a graphically violent, unjustified act, and an ambiguous resolution for the perpetrator. Therefore, we suggest it will promote moral rumination among viewers as they try to internally resolve their emotional and cognitive responses to the narrative. In this narrative, Dr. Melfi is raped on screen. Even though the perpetrator is quickly apprehended, he is also quickly released due to a procedural error. The story line then deals with the fallout from this situation: Dr. Melfi’s son wants to take violent revenge on the perpetrator himself whereas Dr. Melfi’s own therapist advises her to trust in law enforcement for justice. Dr. Melfi is unsure what to do, and ponders if she should confide in her patient and mob boss Tony Soprano to settle the score.

This article is organized as follows. First, section, we address specific viewer characteristics that might promote moral rumination in the particular media scenario we examine. Next, we examine concepts specific to the media realm, such as transportation, which may influence moral rumination processes via involvement in stories and characters. Then, we discuss how viewer responses such as appreciation may be driven by moral rumination. Finally, we introduce a quasi-experimental study examining the role of controversial media in promoting moral rumination.

Viewer characteristics

One of the most important antecedents to moral rumination may be rooted in viewer-specific characteristics such as sensitivity to specific moral issues. For example, Raney and Bryant (2002) found that vigilantism and punitiveness strongly predicted impressions of justice restoration in narrative featuring a violent sexual assault. We predict that for those viewers for whom vigilantism and punitiveness are central concerns, the content and moral conflict in the chosen episode will provoke increased moral rumination due to the focus on these themes in the narrative. Additionally, due to the nature of the content, we also examine relationships between gender and moral rumination.

Hypothesis 1 (H1):

Viewer traits including vigilantism and punitiveness will lead to increased moral rumination in viewers after viewers are exposed to content conflicting with the normative justice sequence normally satisfied via narrative content.

Transportation

Given previous research on the centrality of moral rumination in moral education, we are also concerned with how viewers may be influenced by narrative to change or alter their existing attitudes. The degree to which a narrative is successful in influencing viewers, research shows, is dependent on the degree to which the narrative or characters featured in it is engrossing and immersive, or the degree to which the narrative is able to transport viewers to a fictional world (Green & Brock, 2000). Transportation can be seen as a “flow” experience in constructing mental models of a story that is accompanied by the positioning of oneself in the story world (Busselle & Bilandzic, 2008). Transportation seems to facilitate attitudinal and behavioral effects, among which the endorsement of and belief in story-consistent beliefs are created (Busselle & Bilandzic, 2008; Green & Brock, 2000). Media involvement (or transportation) may shape the interest, attention, and desire of the viewer to spend time thinking about the story world, which would then lead to a greater likelihood to reason within the story narrative or story world as well as deliberate about the story world as if it were real. Therefore, transportation may increase moral rumination via character involvement and liking.

Hypothesis 2 (H2):

Moral rumination will be predicted positively by character liking, victim sympathy, and transportation.

On the other hand, transportation may also involve a subsuming of the viewer’s own moral beliefs into the beliefs of the character or the justification presented in the story realm. Thus, increased moral rumination may occur only if viewers are not so swept up in the story realm; for example, if they strongly agree or disagree with the decision made by the main character. This is a modification to the reasoning presented in H1:

Hypothesis 3 (H3):

The relationship between transportation and moral rumination will be moderated by the extent to which the viewer agrees with the moral decision presented in the show.

Appreciation

Finally, and perhaps most relevant to our story line, Lewis et al. (2014) found that when morality is conflicted in narrative, it can promote deliberative processing in viewers, which leads to greater appreciation. Appreciation is distinct from enjoyment, and includes finding meaning in the content, thinking about the content, and feeling like the content will stay with you a long time (Oliver & Bartsch, 2010). Lee and Shapiro (2014) also linked moral conflict in content to appreciation:

When situations in narratives are complicated and it is difficult to make clear judgments about intentionality, the process of making moral/intentionality judgments and forming affective disposition could be a truth-seeking process. (p. 160)

Therefore, it may be that presenting content that either promotes a moral conflict in the narrative or challenges the morality of the viewer will provoke greater moral rumination in viewers as well as lasting appreciation for the narrative. This would offer a rationale for why moral conflict in content could promote moral education among viewers: The conflict promotes appreciation, lasting thought, and pleasure when thinking about the narrative via the moral rumination the viewer engages in during the episode. Thus, our final hypothesis is:

Hypothesis 4 (H4):

Moral rumination will be positively related to narrative appreciation.

Method

Participants

The sample consisted of 162 undergraduate students from a Dutch university (F = 59.9%, M_age = 19.7, SD = 2.48), who participated in the study to fulfill a class requirement. They were informed that they were taking part in a study on values in fiction, that the show they would watch would include a potentially traumatic sexual scene, and that they could refuse to participate for any reason and still receive credit. Nine participants removed themselves from the sample before viewing the episode at this point. The sample, which was used during analysis, therefore consisted of 153 participants.

Procedure

After filling in a consent form, participants viewed the stimulus material in a cinema-style classroom. Before the screening of the story line, participants filled out a questionnaire including demographic questions, preference for crime drama and previous viewing of The Sopranos, as well as questions regarding vigilantism and punitiveness for rape crimes. Next, participants viewed an edited version of the episode that included only the scenes that made up Dr. Melfi’s story line (duration 25 minutes, 21 seconds). The story line was paused at two times (7 minutes, 34 seconds and 25 minutes, 17 seconds). The first pausing was after the rape scene, when the victim (Dr. Melfi) has the opportunity to tell Tony Soprano about her assault. The video was paused while she deliberates if she should tell Tony or not. At this pause, participants were asked to fill out a second questionnaire including their understanding of the story so far, their transportation into the story line, and their liking of and sympathy toward Dr. Melfi. The scene then continued with the final shot of the episode in which Dr. Melfi tells Tony that she does not want to tell him anything (thereby not choosing violent reprisal for her rapist through Tony). After this scene, the episode and thereby the screening ended and the participants were asked to fill out the last questionnaire, including the extent to which the viewer agreed with Dr. Melfi. This concluding response included an open-ended question option in which the participant could expand on their agreement, a measure of the appropriate punishment for Dr. Melfi’s rapist, and appreciation of the story line.

Measures

Vigilantism

Vigilantism was measured via a 16-item measure (Raney & Bryant, 2002), which tapped into vigilante attitudes about justice (α = .89, M = 3.12, SD = 1.05). Sample items from this scale were “Since too many criminals get off on technicalities, private citizens should bring about justice in situations if they have a chance,” and “No punishment inflicted by a private citizen is ever too severe for a murderer.” The items were rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Punitiveness

The preexposure measure assessed the degree of punitiveness participants felt in punishing a rape perpetrator. Participants read several crime scenarios adopted from Buckholtz et al. (2008) and indicated how appropriate they felt the suggested punishments were. The punishment scale consisted of six items (adapted from Okimoto, Wenzel, & Feather, 2012): “I want John to suffer in some way,” “John should do jail time for his offense,” “John should have some of his rights taken away,” “I think John should be put on parole,” I want to see John punished for his actions,” and “I want to make sure that John is punished at least as severely as his offense demands.” The items were rated on a 7-point scale from 1 (strongly disagree) to 7 (strongly agree). A punitiveness-for-rape measure was created by averaging the rating of the six items for the rape scenario (α = .68, M = 6.06, SD = .83)

Transportation

The degree to which respondents were transported into the narrative was measured with 9-item index adapted from Green and Brock (2000). Sample items for this measure were: “While watching the storyline, I found myself thinking of ways the story could have turned out differently,” and “I was mentally involved in the storyline, while watching it.” The items were rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). A transportation scale was created by averaging the ratings of the nine items (α = .66, M = 5.00, SD = .82).

Victim sympathy

Based on Raney’s (2005) study on punishing media criminals, a single-item victim sympathy measure was used that tapped into empathy felt for Dr. Melfi (“How sorry did you feel for Dr. Melfi?”). The item’s categories were anchored between 1 (not sorry at all) and 7 (extremely sorry) (M = 6.51, SD = .80).

Character liking

The liking felt for Dr. Melfi was measured using six items adapted from Krakowiak and Oliver (2012) (e.g., “I like Dr. Melfi,” and “Dr. Melfi is fascinating”) The items were rated on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) (α = .73, M = 4.71, SD = .84).

Appreciation

At the end of the screening, that is second stop point, we also measured appreciation (Oliver & Bartsch, 2010). Appreciation was measured via two items: “I was moved by this storyline,” and “This storyline will stick with me for a long time.” The items were anchored at 1 (strongly disagree) and 7 (strongly agree). The scores for the two items were combined into a single measure (appreciation: r = .70, M = 3.88, SD =1.43).

Moral agreement

The last question that participants had to answer was whether they were in agreement with the choice made by Dr. Melfi (not to disclose to Tony what had happened) in the final scene. The item asked: “To what degree are you in agreement with the choice made by Dr. Melfi at the end of the storyline?” The item was rated on a 7-point scale ranging from 1 (not at all) to 7 (very much) (M = 4.40, SD = 1.73).

Moral rumination

The moral agreement item was followed up by an open-ended question in which participants were asked to state their reason for (dis)agreeing with Dr. Melfi’s choice. These answers were later recoded for analysis, whereby the answers were coded for moral rumination.

Moral rumination was coded for the complexity of their argumentation, with coding categories: “no answer” (N = 15), “no rumination” (N = 46), “simple moral rumination” (N = 70), and “complex moral rumination” (N = 22). No rumination was a straightforward “She was wrong,” or “I don’t agree.” Simple moral rumination illustrated the reason why participants (dis)agreed with Dr. Melfi’s choice (e.g., “I understand that she does not want to involve her clients in her problems, it is a violation of her professional code”) whereas complex moral rumination showcased a nuanced deliberation including illustration of several sides to the issue (e.g., “I do agree with her choice. It would be terribly unprofessional of her to discuss private matters with a client. Furthermore making yourself judge, jury and executioner is a gross violation of the law. Even though as a viewer you hope for this to happen, in real life this should never occur”). The interrater reliabilities were calculated using the macro by Hayes and Krippendorff (2007), and the interrater reliability for moral rumination was acceptable, α = .79 (Lombard, Snyder-Duch, & Bracken, 2002).

Results

To test our first hypothesis, that vigilantism and punitiveness would promote moral rumination in viewers, we examined the correlation matrix between these variables and moral rumination (Table 1). None of the proposed individual difference variables were directly related to moral rumination (all p > .05); thus, we have a lack of support for H1.

Table 1. Pearson’s correlation between variables.

	2	3	4	5	6	7	8	9
1. Gender	0.08	.31**	.23**	.24**	0.03	0.11	.29**	0.14
2. Vigilantism	–	.38**	0.00	−0.01	−.36**	0.14	0.12	−0.09
3. Punitiveness		–	0.13	.31**	−0.12	.16*	.29**	−0.01
4. Liking			–	.36**	−0.01	.27**	.47**	.18*
5. Sympathy				–	−0.08	.28**	.34**	.18*
6. Agreement					–	−.20*	−0.01	0.11
7. Transportation						–	.59**	.29**
8. Appreciation							–	.24**
9. Rumination								–

*p < .05. **p < .01. ***p < .001.

However, some of these variables were related to other potential predictors of moral rumination. For example, being female was positively related to character liking, character sympathy, and appreciation. Vigilantism was positively related to punitiveness, and negatively related to agreeing with Dr. Melfi’s decision not to tell Tony about her assault. Punitiveness was positively related to victim sympathy and transportation. Therefore, it could be that these variables were exerting a distal influence on moral rumination via the mediators of character liking, victim sympathy, agreement with the story line, and transportation.

To test our second hypothesis, that transportation will be positively related to character liking, victim sympathy, and moral rumination, first the correlation matrix was examined. As expected, transportation was significantly and positively related to all three variables. Next, a regression analysis was run regressing moral rumination on viewer characteristics in the first step, character liking and victim sympathy in the second step, and transportation in the third. The overall model was not significant for the first step, F(3, 146) = 1.23, p = .30. In the second step, the model was significant, F(5, 144) = 1.89, p < .01, R² = .06; however, none of the predictor variables were significantly related to moral rumination (all p > .05). In the third step, the overall model remained significant, F(6, 143) = 3.59, p < .05, R² = .10, and of the predictor variables, only transportation into the story line significantly predicted rumination (β = .28, p < .01). Thus, transportation was predictive of moral rumination, but did not appear to do so via the pathways of victim sympathy and liking. H2 was therefore partially supported.

To test our third hypothesis, that the relationship between transportation and moral rumination will be moderated by the extent to which the viewer agrees with the moral decision presented in the show, a moderation analysis was run using the PROCESS macro (Hayes, 2013; Model 1, 1,000 samples). Results indicated that although the overall model was significant, F(3, 146) = 7.30, p < .001, R² = .12, and both transportation (B = .34, p < .01) and moral agreement (B = .09, p < .01) were related to moral rumination, the interaction between transportation and agreement was not significantly related to moral rumination (B = −.07, p < .13). Thus, H3 was not supported.

To test our fourth hypothesis, that moral rumination would predict appreciation, we first examined the correlation matrix (Table 1). Given the strong positive relationship between moral rumination and appreciation, we found support for H4. However, to more rigorously examine this relationship and control for possible other factors influencing appreciation, the same regression analysis from H2 was conducted, with appreciation regressed on individual difference variables in the first step, character variables in the second, and moral rumination in the third (Table 2). As can be seen in Table 2, punitiveness, character liking, and moral rumination were all significantly and positively related to appreciation.

Table 2. Appreciation regressed on individual difference variables, character variables, and moral rumination.

	Step 1		Step 2		Step 3
	B(SEB)	β	B(SEB)	β	B(SEB)	β
Gender	.60 (.23)	.21**	.30 (.22)	.11	.26 (.22)	.09
Vigilantism	.01 (.11)	.01	.06 (.10)	.05	.07 (.10)	.05
Punitiveness	.37 (.15)	.21**	.24 (.14)	.14*	.26 (.14)	.15*
Liking			.63 (.12)	.37***	.60 (.13)	.35**
Sympathy			.25 (.14)	.13*	.21 (.14)	.11
Rumination					.27 (.12)	.16**
F	6.55***		12.11***		11.23***
R^2	.11		.30		.32

*p < .05. **p < .01.. ***p < .001.

Discussion

The central concern of our study was to examine the role of graphic, unjustified, and unpunished violence in media content with no moral resolution in promoting moral rumination among viewers. Can morally open-ended story lines promote moral rumination and appreciation in viewers? The answer from our results is yes. Moral rumination was predicted by transportation into the story line and was related to increased appreciation for the narrative episode. This supports the notion of media as a “morality sandbox” in which to play with or test out moral concerns, even in regard to controversial, violent, and unresolved conflict. This is in line with Potter and Tomasello (2003), Hall (1993), and others who suggest that going beyond the media effects paradigm to examine viewer responses to violent content fundamentally changes how we can interpret their responses and reactions to this type of content. It may be that moral rumination on the part of the viewer negates or suppresses the possible negative social learning effects postulated in social learning theory (cf. Bandura, 2001). However, our results only begin this investigation and clearly the field must continue in this area to better understand this complex relationship.

In contrast to existing research, our results do not suggest that viewer characteristics, such as punitiveness or vigilantism, play a large or direct role in moral rumination. Therefore, more research is required to understand which individual differences do play a role in moral rumination. Future research could include morality subcultures as defined by professional or personal histories. For example, Van Ommen, Daalmans, Weijers, de Leeuw, and Buijzen (2016) showed that responses to The Sopranos differed greatly for prisoners, law enforcement agents, and civilians. Alternately, Tamborini, Eden, Bowman, Grizzard, and Lachlan (2012) proposed that different moral dimensions are more or less salient for specific viewer groups; therefore, these groups react differently to violent content. Tamborini et al. (2012) based their subcultural groups on moral intuitions, which are gut reactions to moral behavior (cf. Haidt, 2001). In line with Haidt, we suggest that moral rumination may override moral intuitions, particularly when viewers are prompted to ruminate and make arguments supporting their positions as they did in the current study. In this study, we asked participants to reassess their initial reaction by pausing the episode at specific times, thus promoting rumination and rational consideration of the complex moral dilemma presented rather than intuitive response. Similar to Lewis et al. (2014), this reassessment process took time, and we saw a positive relationship between the rumination and appreciation for the episode.

Additionally, character liking and victim sympathy, though related to moral rumination, transportation, and appreciation, seem to be covariates rather than causal factors in promoting moral rumination. Therefore, we suggest that moral rumination may be tied specifically to the conflict presented in the story line, the specific act featured, or the ambiguity of moral resolution provided by the narrative, rather than necessarily being evoked by the feelings viewers have toward the characters. That is, rather than ruminating because we are concerned for the characters, we do so because of moral conflicts or atypical narrative structures presented by the story line itself. This supports the notion of moral rumination via moral conflict put forth by Lewis et al. (2014). In the current study, in which there was limited affective information provided to viewers about Dr. Melfi (beyond that she was female, a psychologist, and a central character), our viewers may have been relying on the rationalist sequence, and waiting to interpret cues in terms of the rapist’s intentionality, or Dr. Melfi’s intentions in telling (or not telling) Tony. The lack of obvious outcome or consequence in the narrative may have promoted conflict in the viewers, but not via the paths of character liking or victim sympathy. Perhaps viewers were ruminating in part because they were not given any obvious narrative or affective cues to help them decide right or wrong in the episode. Future research should vary the amount and type of narrative cues provided to understand what level of narrative uncertainty and conflict provoke ruminative processing.

Limitations

Our study, though supportive of our general premise, was not without limitations. First, the low reliabilities of some of the measured variables are of concern, particularly the transportation variable. This low reliability may have resulted in inflated Type II error, particularly in terms of the individual difference variables. We are not sure why the reliability was so low for this variable, as we used an established scale that has been validated in multiple past studies. On the other hand, we do not believe that the low reliability of transportation is, in the end, detrimental to the most important finding of our article—that specific types of moral content may provoke rumination. It may be that the low reliability for transportation was due to the setting of the viewing room in an existing classroom, which may have limited the ability of the narrative to transport the viewers. Future research should ideally simulate a more naturalistic viewing situation, and use multiple stimuli to examine the processes discussed here.

Next, the moral violation and subsequent justice restoration in the current study may have been overly complex for viewers unfamiliar with the show. In general, we perceived that viewers were presented with a conflict in that Dr. Melfi had two clear choices of justice restoration against her transgressor: via the legal system or via direct retributive action (i.e., telling Tony). However, her choice was also complicated in that she had a professional relationship with Tony, which may be compromised if she utilized his direct retributive methods on her behalf. Indeed, examining the open-ended responses suggests that many viewers used Dr. Melfi’s professional occupation to rationalize her refusal to tell Tony about her assault, as it would damage their professional relationship as doctor and patient. Relatively few viewers indicated they believed that the legal system offered a viable alternative for justice in this specific case. Instead, they perceived her choice as being between retaliation versus no retaliation. Therefore, there are many possible ways in which her decision can be analyzed beyond the simple conflicted choice we presumed.

Further, it may be that viewers with a broader conceptualization of these types of possible outcomes or perspectives are in general are more inclined to moral rumination. Also, viewers with a strong need for cognitive closure may have been particularly concerned with the lack of resolution, which may have led to more rumination as they attempted to make sense of Dr. Melfi’s silence. However, in the current study, we did not look at individual difference variables such as need for cognitive closure, preference for cognition, or levels of moral maturity (as measured with a “defining issues test”; cf. Rest, 1979, 1986) as possible predictor variables for moral rumination. Given the strong reliance of our sample on narrative cues versus intuitionist or affective cues, in future research we may focus more strongly on narrative plot structure and the conflicts perceived in the logic or rationalist sequences versus, for example, character liking and affective feelings about the characters as predictors of moral rumination.

Finally, although our results are suggestive of the notion that viewers are using the media to think about, reflect on, and simulate moral decisions, we did not examine long-term effects or the lasting impression made by the episode. We simply measured perceptions of long-term effect. Future research should examine the role of moral resolution in promoting long-term rumination and processing of these types of stories. In addition, change in moral judgments over time via the processes of moral rumination and appreciation should be examined. Although Eden et al. (2014) suggest that exposure to drama over long periods of time has an effect on moral judgments, it would be helpful to demonstrate that this change comes about via the moral rumination provoked in viewers versus other types of exposure effects such as priming.

Conclusion

In conclusion, the current study gives an exploratory starting look at the role of viewer characteristics and media content in determining moral rumination and subsequent appreciation of television content. Importantly, the study does so using a controversial and graphic episode of unjustified violence, which has since become simply a forerunner of similarly focused episodes. Our findings suggest that rather than act as an agent of desensitization and normalization in viewers, this type of unjustified, violent, and unresolved content may promote moral rumination, reflection, and meaningful thought in viewers if it is paired with an explicit moral conflict and deliberation among the characters involved. Future research should work to disentangle moral conflict from controversy to examine the long- and short-term effects of adding moral complexity to graphic content in narrative.