Cal Boye-Lynn is a Linguistics major at UC Santa Cruz. He is a THI 2023-2024 Undergraduate Research Fellow and the Winner of the Bertha N. Melkonian Prize for his project, “Investigating visual information as a constraint on sound change.” In April, we discussed Boye-Lynn’s project, which explores how small perceptual biases might cause large changes in the sounds of the world’s languages, and the relationship between linguistics research and daily speech.
Hi Cal! Thanks for chatting with us about your THI-supported research project. To begin, could you give us an overview of the project?
Of course! I’m looking at how small perceptual biases might cause large changes in the sounds of the world’s languages. More specifically, I’m trying to figure out why a particular sound change, palatalization, occurs, to get at how we gather and utilize audio and visual information in understanding our conversation partners.
Your project hypothesizes that visual information–what happens on the face of the speaker– plays an important role in certain unidirectional sound changes in language. Could you give an example of the kinds of sounds that interest you and why you believe further investigation of visual information might be necessary to account for these “unidirectional sound changes”?
Stevenson 225, the Perception Lab, where Boye-Lynn is running his speech perception experiments.
Absolutely. To start, a sound change is when a sound or set of sounds in a language changes over time, often over multiple generations of speakers. For instance, English vowels have changed a lot –look up the Great Vowel Shift, if you’re interested! When we talk about unidirectional changes, we mean sound changes which only ever seem to occur in one direction. For example, palatalization, the sound change I’m curious about, takes a word like “beauty” and, over a period of time, turns it into a word like “duty!” However, there really is no “un-palatalization,” where “duty” would become “beauty” again.
While the results are not yet in, it seems like there could be enough auditory information for people to tell them apart, which would be contrary to previous claims. This makes me think that something is actively pushing people to think they heard “duty” when they really heard “beauty.” That’s where faces come in!
When you start the word “beauty,” your lips briefly close then, as you open them, it’s believed that the corners of your lips pull back into a smile-like gesture. When you start the word “duty,” meanwhile, you just get that little smile. It’s possible that seeing the lip corners moving during “beauty” is taken as evidence of a word like “duty.” However, because “duty” has no lip closure, there is no evidence that would push people to see “duty” as “beauty.”
In order to test all of these possible sources of information carefully and systematically, I’m beginning by working with the auditory information alone. Once I’ve got a sense for what I’d be looking for, I’ll look at how this visual information plays a role.
You are currently running two experiments. What are the goals for each and what results have you gathered so far? Are you seeing what you expected to see?
A visualization in Praat, of the final stimuli in the experiment. The top panel is a waveform and the bottom panel is a spectrogram. Notice how clean the vertical lines are in this spectrogram? That’s because this has been resynthesized, taking measurements from the previous spectrograms and using them to create new sounds that tweak certain measurements!
The first experiment is currently running! The second experiment, the follow-up, is contingent on the results of the first, so it will be up once the results from the first one are in.
The first experiment is meant to see if the sounds involved in this change are difficult to tell apart purely because of how we first gather auditory information. However, I suspect that alone won’t explain the change. Pilot data from just myself seems to show that I’m not having any particular difficulty, but I’ve had a lot of practice!
Assuming the rest of the results follow suit, the follow-up experiment could take a number of forms. I’m most interested in following a line of inquiry about the just-noticeable differences between subcomponents of the sounds I’m currently examining, to make sure there are no perceptual biases the first experiment missed.
I’m curious about the relationship between your research and your lived experience. As someone who studies sounds’ perceptual biases, I’m wondering whether you find yourself reflecting on your own or others’ speech habits (including sound and visual cues) in your daily life?
Our speech is full of little hints about not just what we’re trying to say, but who we are and how we want to be perceived.
How we communicate has always been a very personal question for me, and one I’ve asked a lot throughout my life. So, asking questions about how we hear? That’s deeply meaningful to me, and I’m grateful that I have this opportunity.
To more directly answer your question, absolutely, there are always little things in how people talk! Our speech is full of little hints about not just what we’re trying to say, but who we are and how we want to be perceived. It’s fascinating and beautiful! Non-linguists often seem nervous when I point out the cool stuff they do. I think a lot of people think I’m trying to correct them, but I’m really not! Our speech is a kind of quiet, unsung miracle. It’s effortless, for so many people, but its complexity at every level constantly astounds me.
How has your THI Undergraduate Research Fellowship aided you in your research?
The THI Fellowship paid for my insurance. That may sound boring, but it lightened my financial load significantly. It was also an accident! I had sent the school a check, but it had yet to process. That meant I could relax a little more Winter Quarter and focus on this work. The THI gave me a quarter without worrying about insurance bills. Thank you for that!
Finally, what’s your favorite place in Santa Cruz?
That’s a hard one. There are a lot of wonderful places in Santa Cruz itself, but there’s a place up on campus that beats it out for me. There’s a trail that runs from the Cowell-Stevenson area down to the Village, where I lived last year. It’s beautiful. I got to see the seasons change through the trees and undergrowth. As much as my poor feet have been thankful for the reduced step count this past year, I miss walking that trail every day!
Banner Image: One of the original recordings which were manipulated to create the experimental items, visualized using Praat. The top panel, in red, is a waveform, which shows pressure changes over time to create the recorded sound. The bottom panel, in black and white, is a spectrogram, which shows which frequencies in the recorded sound are louder or softer. The x-axis is time and the y-axis is frequency, with darker areas representing when a frequency was louder at a moment in time, and lighter areas representing when a frequency was quieter at a moment in time. Take note of how fuzzy it is, and the shapes of the dark bands (called formants).