“I'm not bad, I'm just drawn that way”: Immoral characters, aesthetic appreciation and sympathy in literary reading

:speech_balloon: Speaker : Julia de Jonge, MSc

:classical_building: Affiliation : Department of Foreign Languages & Literatures at University of Verona; Faculty of Social Sciences at Vrije Universiteit Amsterdam; The Empirical Study of Literature Training Network (ELIT).

Title : “I’m not bad, I’m just drawn that way”: Immoral characters, aesthetic appreciation and sympathy in literary reading.

Abstract :
This paper argues it is not a fictional figure’s moral nature, as intended by the author, that affects the reader’s involvement with the story and the character, rather, readers’ moral judgement. The current experiment shows that while moral judgement of a character is mostly in line with intended character morality (e.g., doctor=good, Nazi=evil), moral judgement is a crucial aspect in reader’s aesthetic evaluation of a narrative and sympathy for its protagonist. Furthermore, an intercultural comparison was made between German and Italian readers showing that moral judgement differs among cultures when it comes to culturally sensitive figures (here, a Nazi-character).


:newspaper: Long abstract

While different reader-responses to narratives with moral vs. immoral characters have been attributed to multiple factors (manipulated character features [Salgaro et al, 2021]; moral disengagement [Krakowiak & Tsay, 2011; Shafer & Raney, 2012]), a causal element still seems to be missing: what causes different readers’ reactions – the constructed morality of the character or the perception of such morality by the readers?

In this study, we compared three different moral characters, intended to be good (doctor), bad (corrupt prison guard), or evil (Nazi-officer). Moreover, perceptions of morality (i.e., moral judgement) likely differ among cultures, in particular when it concerns culturally sensitive characters, such as Nazis or Gestapos. Therefore, an intercultural comparison was also made between German and Italian readers of a text.

Previous research showed that fictional framing leads to a “suspension of moral judgement” (Vaage, 2013, p. 226-7). Because fictional figures don’t have real-life consequences, we would allow bad characters a much higher degree of moral disengagement. However, Salgaro et al. (2021) argue that the German readers in their experiment applied a so-called “reality check” to the Nazi-character in the story, that hindered readers to enjoy the fictional nature or the story itself. Such “perceived realism” seems indeed to play a role in the perception of literary characters (Krakowiak & Oliver, 2012). It is likely that a story presenting a Nazi-protagonist evokes a traumatic past for German readers who, thanks to education and Holocaust studies, might still feel this historic reality as close and related to their actual reality. However, one may wonder whether this would be similar for Italian readers, who had fascism in their historic reality but a different education and a different way of coming to terms with the history of World War II.

The current study aims to replicate and extend this study of Salgaro et al. (2021), using the same manipulated fictional narrative and testing similar concepts, yet including Italian readers. Perceptions of three fictional figures in passages from José Saramago’s Blindness (1996) were compared. The narrative is about a man who becomes blind and relays how the man navigates the world around him with his sudden loss of sight. Four sentences were manipulated to reflect the main character’s morality by referring to past moral actions. The story was manipulated to present either a good, a bad, or an evil protagonist: a doctor (good), a corrupt prison guard (bad), and a Nazi (evil). The experimental modification was limited to the insertion of a few sentences which refer in an indirect way to respectively good, bad, or evil deeds the protagonist had committed many years ago.

Results of our experimental study (N = 202) show that it is not a fictional figure’s moral nature per se that affects the reader’s aesthetic evaluation of the story and sympathetic involvement with the story character, but rather the moral judgement performed by the participants. The doctor-character received the highest moral judgement, the Nazi-character the lowest moral judgement. In line with these results, our readers sympathised the most with characters they considered morally good people, even if the character was intended as immoral (F (1, 197) = 29.056, p < .001, ηp2 = .129). This indicates that the intended character morality itself did not affect sympathy for the character, but rather the readers’ own moral judgement did. Moreover, the story with the Nazi-character was perceived as most aesthetically pleasing (F (2, 197) = 2.702, p = .070, ηp2 = .027), specifically when moral judgement for the Nazi-character was positive (F (1, 197) = 10.085, p = .002, ηp2 = .049).

We did not find any significant differences of character morality or moral judgement on character realism among the Italian readers. It might be that the Italian readers did not experience the “reality-check” as suggested by Salgaro et al. (2021) for German readers regarding the Nazi-character. Overall, scores for character realism are low (Mgood = 2.32, SD = 1.58; Mbad = 2.34, SD = 1.61; Mevil = 2.13, SD = 1.58). This might explain why Italian readers perceived the narrative as more aesthetically pleasant than the German readers. Because the potential “reality-check” did not occur among the Italians, this allows them to apply moral disengagement and enjoy the fictional nature of the narrative.

In all, the current study shows that the reader’s judgement of a character’s morality is crucial in the enjoyment of a fictional narrative and involvement with its protagonist, and that perceived character realism plays an important part in intercultural perspectives on immoral fictional figures. It is therefore important to highlight the divergence of reading experiences and lay out that moral nature is not an absolute value, but rather flexible. Implications for future research are thus that moral disengagement is a fundamental characteristic of literary reading and maybe one of the reasons for its fascination.

Krakowiak, K. M., & Oliver, M. B. (2012). When good characters do bad things: Examining the effect of moral ambiguity on enjoyment. Journal of Communication, 62(1), 117–135. doi:10.1111/j.1460-2466.2011.01618.x.
Krakowiak, K. M. and Tsay, M. (2011). The role of moral disengagement in the enjoyment of real and fictional characters, International Journal of Arts and Technology, 4(1), 90–101. https://doi.org/10.1504/IJART.2011.037772
Salgaro, M., Wagner, V., Menninghaus, W. (2021). A good, a bad, and an evil character: Who renders a novel most enjoyable?. Poetics (101550). doi:10.1016/j.poetic.2021.101550.
Saramago, J. (1996). Cecità. Torino: Einaudi.
Shafer D. M. & Raney, A. A. (2012). Exploring how we enjoy antihero narratives. Journal of Communication, 62(6), 1028-1046. doi: 10.1111/j.1460-2466.2012.01682.x
Vaage, M. B. (2013). Fictional reliefs and reality checks. Screen, 54(2), 218–237. doi:10.1093/screen/hjt004.


Thank you for your interest in my research topic!

If you have any questions, please feel free to ask them here or during the discussion on Saturday.

Thank you for your presentation! I am curious about the length of narratives and how important you think this might be in moral judgment and aesthetic evaluation, as well as how participants’ views might change over the course of a narrative. If I understood correctly, you took all of your measures at the end of the narrative. Do you have any suspicions about how these might change throughout the narrative? I.e. moral judgment for the Nazi might be harsher at the beginning than at the end, or vice versa? Do you see these results as the consequence of a narrative arc, or just one point in that arc?

Dear Claire,
Thank you for your question!
We did indeed take all measures at the end of the narrative. The narrative was approx. 2600 words long and was presented on 7 (half) pages. On four of those pages one manipulated sentence was added to refer to the protagonist’s (im)moral past. The story itself focused on the narrative present. I agree that readers’ perception of the character and aesthetic evaluation of the narrative possibly change over the course of a narrative, but I believe this might be especially true for longer narratives and narratives more focused on (im)moral actions in the story itself. However, this is definitely a very interesting perspective to consider.
Did that answer your question? :slight_smile:

Thank you for your response and the details of the narrative! It was helpful and I appreciate your insights. I think some of our work overlaps in regards to the power of sympathetic framing and morality and I find the relationship between the two so interesting–your project highlights another important facet of this!

Dear Claire,
I definitely see some overlap in our work, and I agree that sympathetic framing will influence perceptions of the character. I realize I haven’t explained it thoroughly, but in the narrative I used the protagonist is struck by a sudden blindness while waiting at the traffic light and follows him navigating how to get back home. It could be very possible that the narrative itself primes the protagonist as a victim, regardless of his immoral past - in line with what you suggested in your second study. Thank you for your very nice presentation as well!

Hi @juliadejonge Thank you for a great presentation! I had a similar question to @clamwood related to the discussion in response to Melissa Seipel’s and @mas29 presentation on Thursday about anti-heroes. (I see links between all of your presentations, and it was really nice to be thinking through a topic guided by related presentations throughout the conference).

What is your stance Julia on anti-heroes versus immoral characters? Are they the same or do they differ slightly? And in how far is the development of such a complex character, and whether or not they are protagonist or “side character” influence readers’ moral and aesthetic judgment?

Maybe to take the discussion a bit further and involve @clamwood and @mas29: how are readers’ moral and aesthetic judgements of immoral characters/anti-heroes influenced by the process of character comprehension (i.e., wanting to figure out and understand why a character does and/or says certain things)? And do we need experiments with longer stories (e.g., tv series or novels) to fully investigate this relationship?

I think I am having trouble answering this because it’s complicated. Also, because I am about to suggest that the category of antiheros we used in our study is useful, but not the only way we should be looking at this. Several people have asked if we could say the same things we found in our study about any complex character. In general we would agree.
Image two characters.
One (character A) manipulates and hurts a partner, but sometimes does nice things—even selfless heroic things.
Her partner (B) never does anything heroic, but tolerates A’s hurtful behavior because of B’s long history of neglect, etc. (fill in whatever backstory you want).
My point is that A might fit a definition of antihero and B does not, but audiences are likely to use many of the same strategies to judge, form allegiances with etc. that Melissa and I identified in our study.
Characters can be complex in many ways. Even morality can manifest itself in many ways.
I would say that we should probably pay more attention to the process audiences use, to identify the dimensions and mental strategies that audiences use to make sense of complex characters, than to sort them into categories.
Interestingly, I find myself more likely to form allegiances with character A than B—but that might be just me. I suspect romance fans might have the opposite reaction. Is the concept of partial allegiances useful for a variety of complex characters, not just those who fit a definition of anti-hero.
Also, the morality of a complex character often changes over time. Characters can also be unreliable narrators. In her book Sunshine points of that many readers of Lolita are initially taken in by Humbert (she admits she was) but over time most readers realize the depth of his depravity.
Sometimes our categories do more to conceal than reveal. I think the variable approach—like the one Julia is using is important. But as many authors have pointed out, communication is a process. The process we use to understand characters may need more attention.
I hope I’ve been at least somewhat clear. I’m interested in what you think.

Dear @moniek.kuijpers, Dear @mas29,

Thank for your question Moniek!

I agree with Mike, immorality is complex concept. What one person perceives as immoral, is not necessarily immoral to someone else.
An example is in the comparison between the “bad” and “evil” character in the study/studies explained in my video. While the readers in the study by Salgaro et al. did perceive a difference between the two immoral characters, the readers of our replication study did not perceive them any differently.

Perhaps anti-heroes are immoral character portrayed in a more sympathetic frame. This could be an interesting research question, actually! Either in the process audiences use, like Mike suggests, or in testing different framing of an immoral character (e.g., sympathy vs. apathy from other characters, main character vs. side character, background story vs. no background story).

One thing I didn’t mention from our study is that between character relationships were important. A few things that emerged.

  • Contrasting antihero with even worse characters made anti seem more moral
  • Contrasting antihero with heros
  • Alliances of antihero with heros (Loki and Thor).
  • Sympathetic other characters (particularly family)
  • Liking for antihero (and other judgments) by neutral but supporting characters (Walter White’s wife (Who BTW many viewers disliked but understood as more moral).

Not sure how to fit all that in a dimensional structure.