Instru(mental)

#18 - Music as Moral Evidence

• Brea Murakami • Episode 18

Share episode

Can music change our moral judgments of others? In this episode, we explore how knowing about someone's, or something's, musicality shifts our moral decision-making. This episode's research study includes four related experiments involving musical monkeys, anti-musical humans, dogs, babies, and robots. Then, we explore what the converging results suggest about music’s impact to shift social attitudes and advocacy efforts related to groups that are often dehumanized, like the homeless community or incarcerated individuals. 


References

  • Agrawal, T., Rottman, J., & Schachner, A. (2023). How musicality changes moral consideration: People judge musical entities as more wrong to harm. Psychology of Music, 51(1), 316-336. https://doi.org/10.1177/03057356221096507
  • Robison, M., Aderhalden, F.P., & Joiner, T.E. (2024). Dehumanization and the association with nonsuicidal self-injury and suicidal ideation in an incarcerated population. Crisis, 45(4), 287-293. https://doi.org/10.1027/0227-5910/a000952
  • Schroeder, J., & Epley, N. (2020). Demeaning: Dehumanizing others by minimizing the importance of their psychological needs. Journal of Personality and Social Psychology, 119(4), 765-791. https://doi.org/10.1037/pspa0000199


Resources

00:01: Brea: Hey there, it’s Brea. As a music therapist, there is this movie trope that I see everywhere, from horror movies to zombie movies, and even Pixar animated films. And once you hear about it, I think that you’ll start seeing it everywhere too. Take the movie WALL-E for example. When loveable, compactor cleaning robot WALL-E brings futuristic robot EVE back to his hideaway, he starts handing her examples of treasures he’s found during his cleaning trips, but he kind of freaks out when EVE pulls the magnetic tape out of a VHS of the musical Hello Dolly! When he plays the VHS back to check that it’s okay, it starts playing a song from the musical and WALL-E dances along, using a little hub cap as a makeshift hat.

00:46: Or, (spoiler alert ahead!) in the 2016 zombie movie Train to Busan, the ending scene is of two characters, a woman and a little girl walking alone into a tunnel toward what they hope is safety. We see snipers at the other end of the tunnel take aim at the two figures because the soldiers aren’t visually about to verify whether the woman and girl are infected. Just as the sniper is about to pull the trigger, he hears the little girl start to sing. The soldiers immediately get up to help the two survivors into the safety of the quarantine zone.

01:23: For me, one of the most obvious examples of this trope is from the 2022 horror/sci-fi movie M3GAN (spelled with the number 3) in which a woman creates a life-like robot companion for her orphaned niece, Cady, who is 8 years old. I’m going to play you a scene, but in this scene, Cady is grieving the death of her parents, whom she’s worried about forgetting. Trying to make Cady feel better, the Megan robot audio records a memory Cady just described and then M3GAN begins to sing… 

01:57: M3GAN: Ok, so that’s a memory you’ll never forget.

Cady: What do you mean?

M3GAN: I mean I’m keeping it for you, here.

Cady: One time, she found a cockroach in my school bag. She was upset because I didn’t eat my sandwiches. And then, all of a sudden, this thing crawls up her wrist and she just started screaming like a maniac and ran out of the house. That was pretty funny.

M3GAN: Anytime you want to tell me something special about your parents, something funny or sad or anything at all, you just tell me and I’ll keep it safe and we can listen to it whenever we want. If you should feel alone or that your world has come apart. Just reach out and you’ll see a friend is never very far. Tell me your dreams, I will dream them too. I so glad I finally found you.

03:08: Brea: To me, M3GAN’s robotic voice is just really uncanny valley, and I think watching the clip enhances this effect even more.  If you do want to check out any of these movie clips for yourself, I have linked all three in the episode description.

03:24: In all three of these movie clips, there’s this association between a character who makes music equaling the character being viewed as more humanlike. And it’s in dozens of other movies and TV shows. Clearly, filmmakers are using music whether it’s singing or dancing along, or even just appreciating music as a storytelling shortcut to get us as the audience to feel more empathetic towards these characters and robots and sometimes even the movie’s villain. But, is this storytelling shortcut backed up by empirical evidence? Does music lead us to view others with more humanity? And, can knowing that someone is musical change our moral judgments of them? The research suggests that it’s possible, so keep listening to find out more…

04:28: Of course, the prosocial effects of making music with others is well-documented. In fact, the second-ever episode of Instru(mental) goes over some of the research that suggests that when we entrain with someone else, whether through singing together, or improvising music together, that these shared musical experiences are like a shortcut to entering into a space where we have the same goals as others. Essentially, to create music that sounds good and expresses something. And having this shared musical goal has additional benefits that we trust others, we like others, and we’re more helpful and cooperative with the people we make music with. And also, in episode 17 of Instru(mental), I also talked about how I used group singing to manipulate my wedding guests with this same principle. 

05:20: In my mind, it’s pretty well established that making music with others increases our prosocial feelings toward others. But, to be perfectly honest, the movie trope I just described where simply observing a character making music imbues more humanity onto that  character wasn’t something I noticed until I attended a conference presentation. I first heard about the research we’re covering in today’s episode at the SMPC, which stands for the Society for Music Perception and Cognition conference when it was held in New York City in 2019. At the time, those authors were sharing their preliminary results, and I was super excited to see that they formally published their results in 2023.

06:06: Those researchers were Tanushree Agrawal, Joshua Rottman, and Adena Schachner and their paper in the Psychology of Music journal actually outlines four separate experiments they conducted about whether musicality impacts our moral judgments. Or, put another way, does knowing that a person or an animal is capable of making music, does understanding their musicality change how we treat them? Their experiment didn’t involve actually hearing music or making music with the proposed organisms, the study just included written descriptions of what these organisms were capable of. 

06:46:  Here’s how the basic experimental procedure went. In the researchers’ first study, they recruited 100 adults via an online platform to participate. These 100 adults were presented with a pair of two characters out of nine character possibilities. When they saw each pair of characters, the participants were given the prompt, “if you were forced to harm one of these characters, which one would it be more painful for you to harm?” And, from what I can tell, there wasn’t really any kind of example of the kind of harm that would theoretically be imposed on each character. The participants just had to choose which of the two they were presented with, which one would be more difficult for them to hurt in an abstract kind of way.

07:32: This prompt was how the researchers measured the moral worth of each character. Theoretically, it’s harder for us to imagine inflicting harm or pain onto others that we deem to have higher moral standing. So, if Character A was consistently chosen as more difficult or more wrong for a participant to harm above most other characters, then  Character A would be seen as more morally worthy of protection. This “wrongness to harm” prompt has been used in previous research about moral standing, and generally, as an example people rate plants as least wrong to harm, followed by fish being slightly wrong to harm, followed by apes as being more painful to harm, finishing up with human infants being the most morally wrong to harm.

08:25: And, these findings seem pretty intuitive to me. In my estimate, plants don’t have the same capacity to feel emotions or have consciousness, but these mental and emotional capabilities are higher in fish or apes, and of course a human infant is going to be hardest to hurt because it has the most humanity, literally.

08:48: But, in this study, who were the nine characters that the first set of participants were judging? The characters ranged from a 5 month old baby, to a robot, a pet dog, a green frog, two different monkeys, two different adult men, or the participant themselves. A picture and a short description of each character was included with each pairing, and it was these descriptors that subtly indicated to participants that some characters had musicality, and others didn’t mention anything about music.

09:25:  For example, between the two monkey characters, there was Gabe, a capuchin monkey who listens when his zookeeper is playing music and Gabe dances along by bobbing his head, contrasted with Toby, a six year old capuchin monkey, who lives in a large enclosure at the Bronx Zoo in New York City. For the musical human character there was Eric, a thirty-year-old musician and devoted music-lover living in New York whose non-musical counterpart was Todd, a thirty two year old accountant and devoted trivia player living in Chicago. Throughout the first experiment, all 100 participants rated every possible pairing of the 9 characters, resulting in 36 pairwise comparisons in who was more painful for the participants to harm. 

10:16: This first study’s results overall found that the 5 month old baby was almost always more painful to harm, with the second most painful character to harm being the pet dog. The third and fourth most difficult characters to harm were actually the musical human and the musical monkey, which were both rated as having higher moral standing than  the non-musical human and the non-musical monkey. There were also statistically significant differences when comparing the pairs of monkeys and humans. Specifically, the musical monkey was rated as more painful to harm 75% of the time, compared to its non-musical equivalent. While 60% of participants rated the musical human as more difficult to harm than the non-musical human. From the results of this first study, it does seem as though the researchers’ hypothesis is on track, there may be something about a musical character that makes us feel differently toward them.

11:30: The researchers did a follow up study with a very similar procedure, but with a few subtle changes. First, they recruited 150 adults for this second experiment to verify the reliability of the first study’s results comparing the same pairs of 9 characters. But, rather than individually named monkeys like Gabe or Toby being compared, the researchers wanted to see if participants rated an entire species of monkeys differently. So, in experiment 2 the musical monkeys were described as “Calusan monkeys [which are] are native to Central and South America. Calusan monkeys are known to respond to music, by dancing along and bobbing their heads.” This is compared to the non-musical monkey species who were described as, “Sequel monkeys living in tropical rain forests in the Amazon. Sequel monkeys are brown, grey, or white, with lighter fur around their face and eyes.”

12:29 :What did the results of Experiment 2 find? Actually, they were pretty similar results. Once again, the baby was always the most painful to harm, followed by the dog as being second most painful to harm. And, again, the musical versions between the human-monkey pairings were always statistically harder to harm with 65% of participants saying the musical monkey species was more painful form them to harm than the non-musical monkeys, with 63% of participants rated the musical human as more painful to harm than his non-musical counterpart. Experiment 2 had similar results to Experiment 1, which together suggest that if a character is musical, then we have a harder time harming them.

13:18: Like most scientists, these results made the researchers even more curious so they decided to conduct a third experiment. This time, the researchers threw in a 10th character who was a human who was specifically not musical. This anti-musical character was “Matt, a 36 year old who has never enjoyed listening to music, and he doesn’t understand why others seem to find music so appealing.” Matt joined the other two human characters from the first two experiments. Again those were music-positive Eric, a professional musician who enjoys listening to music in his free time and music-neutral Todd, a professional who works a full time job in Chicago. So to recap, in this third experiment, the researchers had a new set of 150 adults rate all pairwise comparisons of these three human characters, along with the musical monkey species and the non-musical monkey species, and a host of other characters like a robot, a frog, a baby, a dog, and the participant themselves.

14:26: What happens when you add the anti-musical character into the mix? Actually, pretty similar results. The musical monkeys were harder to harm than their non-musical counterparts. And, while all three human characters were rated as more difficult to harm than both monkey species, the musical human was rated as more morally worthy than the neutral human, who in turn was rated as more morally worthy of being protected from harm than the anti-musical human.

14:58:  I know that all of these results might be getting a little repetitive over this  podcast, so I really recommend that if you’re curious to check out the full version of the research article which is open access, meaning that it’s not behind a paywall. The graphs for all of these character comparisons are really clearly laid out, with little character icons right there on the graph. They’re actually kind of cute, so I do encourage you to check out the episode description for a direct link to the article where you can review the graphs for yourself. But, just to make sure we’re all on the same page, the evidence from these three studies is really starting to converge in a way that suggests that we are more reluctant to harm characters that have musicality, compared to their neutral or even anti-musical counterparts.

15:49: But, these three studies’ results don’t really tell us why musicality impacted the participants’ moral decision-making about whom they would rather harm when given a choice. The researchers decided to conduct a fourth and final experiment, replicating these pairwise comparisons, but this time asking participants to rate the ten characters not just on which one they’d have a harder time harming, but also asking participants to rank each character’s musicality, intelligence, and capacity for feeling emotions. The researchers hypothesized that people see a character’s musicality as a proxy or indirect measure for the character’s broader emotional and mental abilities. If a character has more psychological depth or agency to make conscious decisions (as inferred by their musical capabilities ), then that character is also more wrong to harm, and thus deserves to have higher moral worthiness.

16:54: To investigate whether intelligence or emotional capacity was a mediator between a character’s musicality and participants’ moral decisionmaking, the researchers recruited another 150 adults, but this time they were undergraduates at a college and had them compare the same 10 characters as in Experiment 3. When you look at the wrongness to harm results, they’re pretty similar to the first three experiments. The musical versions of characters were always rated as more wrong to harm than their non-musical counterparts. 

17:29: But, in this fourth experiment, data were also taken on characters’ musicality, intelligence, and emotional capacity. In terms of musicality, it’s kind of straightforward. The musical human was rated as having more musicality than the neutral and anti-musical humans, and participants similarly rated the musical monkey species as being more musical than the non-musical monkey species. Pretty intuitive, maybe obvious, but when you’re doing science, you do need to verify the things that are seemingly obvious.

18:05: Moving onto the other characters, the musicality ratings for the baby and the dog were rated in the bottom or middle, even though in all four experiments these two characters were always the most difficult to harm, even higher than the musical human or often the participant themselves. So, it’s not just someone or something’s capacity to make music that changes our moral judgment. 

18:32: Looking at all 10 characters’ intelligence ratings, there were some interesting results. Consistently, the musical human character was rated as having the highest intelligence, even over the participants themselves, and the neutral human was rated as more intelligent than the anti-musical human. And, while all humans were rated as more intelligent than both monkey species, the musical monkey species was rated as more intelligent than the other monkey species. And side note here, the robot was rated as more intelligent than the dog, the baby, and the non-musical monkey species, even though the robot was the easiest or the least wrong to harm across all four studies. Which brings us to the final ratings for the characters, which had participants estimate which characters were the most  emotionally sensitive. Once again, the musical versions of the humans or the musical versions of the monkeys were rated as having a higher capacity for emotions than their non-musical counterparts.

19:38: I know this was a lot of data to get through, but I really respect that these researchers decided to conduct three follow-up studies to the first experiment to really suss out whether the trends they found were reliable and to test out different hypotheses for why they were seeing these trends. Bringing all four experiments’ results together, there is pretty strong evidence that when we know that a character is a musical being, that we judge them as being more difficult or wrong to harm. And, there’s something unique about musicality that is influencing our moral judgments, that can’t be explained by emotional capacity or intelligence alone.

20:18: In the discussion section of the research article, the researchers write, “People appear to intrinsically value musicality, in a way not explained by experience or intelligence alone.” It could be that when we know someone or something is musical, that we imagine them to have a richer inner mental life, which in turn makes it more painful to imagine inflicting harm on them. However, the researchers caution that all four experiments’ results should be interpreted with caution and that there are many factors beyond musicality that we weigh when making moral judgments. 

21:11: There are some interesting implications to these findings. The researchers themselves propose that these findings could be applied to animal conservation efforts. As an example, the Save the Whales movement in the 1970s that led to policies that protected whales may have picked up additional momentum when scientists used whalesong research as evidence that whales were sentient beings who were morally wrong to harm. When members of the public had some sense that whales’ communication was music-like, then they might have been more likely to donate time or money on efforts that were trying to prevent harming whales on a global scale.

21:51: As a music therapist, I also see these results as having implications for how we can change social attitudes about groups of people who may be dehumanized, treated as less-than, or just generally segregated away from parts of larger society. For example, many parts of the United States are having issues with people living on the streets due to substance abuse, mental health issues, poverty, or even just having a series of life circumstances happen without having a safety net to fall back on. These folks can be stigmatized and treated really poorly as a result, which likely exacerbates these issues further.

22:33: I found this really stark research study from 2020 that examined how dehumanizing beliefs can lead members of the public to vastly underestimate the need for love and meaning that the people living in poverty or who are unhoused have. Specifically, charitable donors who have good intentions and want to help others in poverty will rate the recipients’ need to eat food as the most important, while rating the recipients’ need to feel loved as least important. In the study I read, these charitable donors’ skewed beliefs led them to donate food-based gifts, when the people they were actually trying to help almost never ranked a gift involving food to be their top priority or  ideal. Instead, the gift recipients said cash or cash-equivalent gifts like a gas card would be ideal and make a meaningful change in their life. It seems to me that even charitable people with good intentions can subconsciously dismiss the humanity of the folks they’re trying to help, which may lead to donations that aren’t as impactful as they could be.

23:43: And, it’s not just the general public’s beliefs that can be impacted, these dehumanizing beliefs can have heavy effects on stigmatized individuals’ mental health. For example, people who are incarcerated can be dehumanized when they have limited freedom and agency when they’re completing a prison sentence. Research suggests that when people who are incarcerated perceive that the officers view them as “just a number,” “stupid,” or “like they’re animals” or otherwise as less-than-human, that these beliefs are associated with higher rates of self-injury and suicide ideation among incarcerated people.

24:45:  Although it’s not super common for music therapists to work with the homeless community or in correctional settings, these are still settings that I believe have so much potential to meet these higher-level, but often ignored needs. Everyone deserves to feel like they have dignity, that they’re capable of creating something aesthetically beautiful, that they have meaning. And, music can be a way for these human-centered needs to be fulfilled.

24:53: Music could also have positive ripple effects that staff could benefit from when observing the folks they work with making music. Expanding out even further, I wonder what perspectives could be shifted among the general public if they could see people they might otherwise avoid doing musicking things. Could seeing members of the homeless community or people in prisons making music, singing, playing drums, dancing, could these types of experiences allow neighbors, and staff members, and policy makers to see  these folks’ humanity with a fresh pair of eyes?

25:29:  Of course, homelessness, incarceration, poverty and any other social attitudes around these issues that we might be trying to change are complex and not easily shifted with seeing or participating in one musical activity, one time. But, these four experiments’ results suggest that our moral judgments and aesthetic judgements might be more closely intertwined than we give credence to.

25:57: The researchers even mention this possibility in their discussion, writing that our association between musicality and intelligence may be further influenced by what genre of music we see someone making. Like, if we see someone playing classical music we may imbue a greater sense of intelligence or sophistication onto them than if we were to see them play pop music. What the music sounds like when we’re observing someone create music may have implications on our moral thoughts about them.

26:30: If you’re looking to examine your own social attitudes, you can of course use the takeaways from this research to pause the next time you notice yourself othering a stranger. It sounds kind of silly, but everyone has sung or moved to music or made music at least once in their life. Music is just something humans do. Imagining the musical version of someone you’re noticing a bias against might be one way to reframe what they’re capable of connecting with. As a start, I’ve linked some video examples including a clip from the 60 Minutes show about Nathanial Ayers, whom the movie The Soloist is based on and a few other feature clips about music therapy in forensic settings.

27:14: Altogether, it’s pretty amazing that not just hearing music that someone else makes, but simply knowing or remembering that people have the ability to make music can shift our viewpoints of other people to be more humane and generous. This really highlights that music is not just a superfluous, aesthetic sense that we have. Knowing someone or something else is capable of musicking might prompt us to reconnect with something intrinsic to ourselves as humans.

27:45: Thanks for taking a listen to this episode of Instru(mental). As always, you can find the research references, movie clips, and resources mentioned in this episode in the episode description. And for fun, if you can think of another movie or TV clip where a character’s musicality makes them more human-like or enhances their humanity, feel free to drop it in a comment. I’d love to see what other examples we come up with. And, while you’re at it, make sure you’re subscribed so you don’t miss the next episode, which will be out soon. 


People on this episode

Head Shot

Brea Murakami

Host