Thanks to the If Books Could Kill podcast and its co-host, Peter Shamshiri, I started listening to 5-4 — “a podcast about how much the Supreme Court sucks” which Peter also co-hosts. It’s good, infuriating, and informative. But handing you a new podcast to listen to is not why I’ve asked you here.
There is a promotion at the half-way point for a newsletter — Balls and Strikes — that another co-host, Michael Morbius, narrates. They seem to run it each episode and I’ve started noticing something. Actually, I’ve noticed that I’ve noticed something. It’s a bit meta.
I’m getting there.
At one point as he speaks I could tell that he starts smiling. The change is clear but undescribable. I don’t know why my brain has picked up on this. Less so, do I know why he’s smiling. So I went to the Internet, as I do, to find out why my brain does what it does.
My first stop was this article in Discover Magazine that showcases a study suggesting that if you can sense a smile in a voice you’re hearing, and not someone you can see, you tend to smile back. The article and the study it links to–well done, consumer science journalism–discuss the lack of research on what constitutes this auditory smile. Checking the paper’s sources, I ended up here: “The vocal communication of different types of smile” in Speech Communication. The study is from 2008 and I’m not sure if I’m going to see what builds upon this research. But I was still curious to see who else out there was wondering, “did I just hear you smile?”
“Smiling voices maintain [increased trust] even in the face of behavioral evidence of untrustworthiness.” (1)
…and here:
“We present an experiment in which participants played a trust game with a virtual agent that expressed emotion through its voice, in a manner congruent or incongruent with its behavior.” (1)
…and here:
“Using an investment game paradigm, we found that positive vocal emotional expression – smiling voice – increases participants’ implicit trust attributions to virtual agents, compared with when agents speak with an emotionally neutral voice. As previously observed, the monetary returns of the agent also affected implicit trust, so that participants invested more money in the agent that was behaving generously.”(1)
And this is the point where I’ve saved the citation in Paperpile, sat back with my arms folded and leaned over to look down into the murky depths of this rabbit hole. I still don’t know what stimuli my brain is picking up that translates into “smile” after Michael says “Supreme Court sucks”, but I can pick up the danger of being able to simulate this in such a way that creates trust between yourself and stranger on the phone.
This is more than just Cash Green’s white voice in Sorry to Bother You, this is the “right voice,” the one that flicks an unknown switch in your head and you picture a reassuring smile. The “right voice” is built upon the research that pull the secrets out of our brains and tools them for algorithmic benefit. The “right voice” won’t just relieve people of their hard-earned money, it will lead them astray, down paths not yet cut.
What do I do? This digression has made me thoughtful. Sigh.
(1) Torre, Ilaria, et al. “If Your Device Could Smile: People Trust Happy-Sounding Artificial Agents More.” Computers in Human Behavior, vol. 105, Apr. 2020, p. 106215. https://doi.org/10.1016/j.chb.2019.106215