Regarding the Robot Perspective: The Effects of Human Bridge Characters on Empathy and Altruism of Readers for Robots

:speech_balloon: Speaker: Katja Warstat-Willms @KatjaWW

:classical_building: Affiliation: RWTH Aachen University

Title: Regarding the Robot Perspective: The Effects of Human Bridge Characters on Empathy and Altruism of Readers for Robots

Abstract (long version below): Human bridge characters can help to circumvent the anthropomorphism of non-human agents in favor of an understanding of fictional narratives through a human perspective (James 2019). A human “bridge” furthermore might positively affect the empathy and altruism for non-human agents. The approach puts into focus the impact sci-fi literature might have on the successful implementation of robots in our everyday lives.
My pilot study investigates the impact of narratives presenting a robot perspective in contrast to the effects of a human “bridge” character on readers’ felt empathy and their altruistic intents, and the degree to which they anthropomorphize robot characters.


:scroll:
Bridge Character Study_Warstat-Willms.pdf (1.0 MB)


:newspaper: Long abstract

Human bridge characters can be a means of circumventing the inevitable anthropomorphism of non-human agents in favor of an anthropocentric understanding of fictional narratives through a human perspective (James 2019, 582). James’ expansion of Keen’s model of narrative empathy (e.g. 2007; 2011) is supposed to account for the positive impact stories revolving around non-human animals can have on our behavior regarding non-human animals in general (James 594-5), a thought also supported by Herman’s idea that “narrative affords a bridge between the human and the nonhuman” (2011, 159). The idea that human bridge characters might not only be used to make us care about non-human animals, but also about non-animal agents like robots, gains importance in face of the ever-increasing implementation of artificial intelligent agents in our everyday lives. Fictional human “bridges”, providing a relatable perspective, might indeed promote empathy for fictional robots and consecutive altruistic behavior for real-life robots. Former studies on related topics have already focused on the effect of narratives of animal suffering (see, for example, Malecki et al.), but a focus on robots may provide new insights regarding the significance of (science-fiction) stories for our real-life human-robot interaction.
My pilot study investigates the impact of a direct robot perspective in contrast to the effects of a human “bridge” in fictional narratives on readers’ felt empathy and their altruistic intents, as well as the degree to which they anthropomorphized robot characters. The following hypotheses were tested:

• H1: Reading a text including a human bridge character will result in higher scores of altruism for robots than reading a text with a robot focalizer.
• H2: Reading a text including a robot focalizer will result in higher scores of anthropomorphism of robots than reading a text with a human bridge character.
• H3: Reading a text including a robot focalizer will result in higher scores of empathy for robots than reading a text with a human bridge character.

Participants of my study were presented with two different texts each (an excerpt of Ishiguro’s novel Klara and the Sun (2021) and one from a short story taken from Winterson’s The Stone Gods (2007)). Group 1 and 2 received texts with a robot focalizer including no human bridge character or texts exclusively featuring human bridge characters, respectively (one original and one manipulation); group 3 and 4 were conducted to control for the effectiveness of the manipulation of my stimulus material and received only the original texts (one with and one without a bridge) or the manipulated texts, respectively. Consecutively, all participants were presented with a questionnaire including questions regarding altruism and their altruistic disposition (Manzur & Olavarrieta 2021), anthropomorphism (Eyssel & Kuchenbrandt 2012), and empathy as well as the empathic disposition (Kuijpers et al. 2014). All items were presented with 6-point Likert scales, ranging from very much (6) to not at all (1), to allow for a measure of degree.

Results were analyzed using pairwise t-tests; in addition, an ANOVA was conducted to test for any interaction between empathy and altruism, because these are, in theory, connected concepts (see above). To account for the possibility that some readers might not tend to empathize with fictional characters at all or do dislike science-fiction and were therefore put off by the stimulus material, an open question at the end of the questionnaire asked how readers liked the excerpts and how they feel about these kinds of stories in general.

The results of the quantitative data showed that the manipulation of the texts (to include or not include human bridge characters) was effective (there were no significant differences between groups 3 and 4), which means that a future study might be conducted based on one literary text and its manipulated form for greater comparability. Surprisingly, though a tendency for differences between groups 1 and 2 was found, these differences went into the opposite direction of that indicated by hypotheses 1 and 3. Altruism for robots was apparently more affected by the group that read texts including a robot focalizer, while empathy for robots was reported as stronger for the group that read texts from a human perspective. In regard to anthropomorphism, no impact of any text could be established. However, the qualitative data indicated that these results might have been affected by the perceived agency (or lack thereof) of one robot character and the resulting decisive preference of one text over the other.

References
Eyssel, Friederike, and Dieta Kuchenbrandt. “Social Categorization of Social Robots: Anthropomorphism as a Function of Robot Group Membership.” British Journal of Social Psychology, vol. 51, no. 4, 2012, pp. 724-31.

Herman, David. “Storyworld/Umwelt: Nonhuman Experiences in Graphic Narratives.” SubStance, vol. 40, no. 1, 2011, pp. 156-81.

Ishiguro, Kazuo. Klara and the Sun. Faber & Faber, 2021.

James, Erin. “Nonhuman Fictional Characters and the Empathy-Altruism Hypothesis.” Poetics Today, vol. 40, no. 3, 2019, pp. 579-96.

Keen, Suzanne. “Fast Tracks to Narrative Empathy: Anthropomorphism and Dehumanization in Graphic Narratives.” SubStance, vol. 40, no. 1, 2011, pp. 135-55.

Keen, Suzanne. Empathy and the Novel. Oxford UP on Demand, 2007.

Kuijpers, Moniek, et al. “Exploring Absorbing Reading Experiences: Developing and Validating a Self- Report Scale to Measure Story World Absorption.” Scientific Study of Literature, vol. 4, no. 1, 2014, pp. 89-122.

Manzur, Enrique, and Sergio Olavarrieta. “The 9-SRA Scale: A Simplified 9-Items Version of the SRA Scale to Assess Altruism.” Sustainability, vol. 13, 2021.

Winterson, Jeanette. The Stone Gods. Hamish Hamilton, 2007.

Cool subject, probably will get more and more relevant by the day. A phenomenon that might play a role in this, I am not sure, is the uncanny valley effect. What’s your idea about that? Uncanny valley - Wikipedia

Also, maybe this work may interest you: Mara, M., & Appel, M. (2015). Science fiction reduces the eeriness of android robots: A field experiment. Computers in Human Behavior , 48 , 156-162. There are more publications of Appel and colleagues that you could look into, if this article seems useful to you.