"Linguistics has a lot of expectations associated with it," says Shiva Upadhye, a third-year grad student in the UC Irvine Department of Language Science. "There are certain theoretical traditions that people commit to. There are certain lines you draw."

UCI's pioneering language science program — the first of its kind in the country — caught Upadhye's attention when she was applying for grad school back in 2021. "Traditionally, linguistics has this certain set of methodologies associated with it, whereas language science is kind of an attempt to break away and just widen the scope," she says. "It's sort of looking out to other disciplines, such as cognitive science, psychology, neuroscience, computational work. It's really kind of pulling from all those different disciplines and applying that to the study of language."

The broader, more flexible approach of language science, still a relatively rare offering among graduate programs, resonated with Upadhye, who holds a UCSD bachelor's degree in cognitive science. "I was actually applying mostly to cognitive science programs" she says. "Because I was coming in from a more interdisciplinary field, I was very drawn to the [language science] program. It was much more in line with the way that I wanted to look at language, which is not a capacity or faculty or even a field of study that's divorced from the rest of cognition."

Upadhye's focus is language production, the process by which a thought is verbalized or converted into text. "I'm specifically interested in looking at production as a form of action," she says. Strangely enough, there's not much previous research to draw upon in this area, despite the fact that it's a crucial part of understanding language use. "Even though language production is a form of behavior that leverages a lot of the same faculties as other forms of action, usually the way that it's been studied has been quite divorced from the rest of action or other kinds of behaviors," says Upadhye.

She's most interested in researching planning in language production — that is, how our brains form what ultimately comes out of our mouths. It also has to do with why we sometimes misspeak, and how we recover from errors to salvage communication. "If I were giving someone these directions, and I ended up saying something like 'turn left' instead of 'turn right,' it's a fairly simple error. But it is fundamentally distorting the message that you want to convey," she says.

She posits that humans have an innate tendency to conserve cognitive resources, which may lead to these types of brain hiccups. "Humans are fundamentally lazy and want to withhold rather than put in effort. And I think that shapes a lot of behavior," she says. "Given that you're operating under cognitive constraints, how do you allocate appropriately at various stages of the production process to basically ensure that you're able to achieve your communicative goals? You want to be efficient, and you want to produce things in a timely manner. But at the same time, you have constraints on processing, so you're not actually able to allocate as much attention or resources to coming up with, let's say, the most accurate choice. And so how do you settle on something that balances utility and effort?"

The flexibility of the language science department has given Upadhye the latitude to explore these types of questions. "A year ago, when I was doing the literature review for another milestone, I sort of felt like, 'Am I straying too far from what is considered traditional or standard in terms of linguistic theory?' And then my advisor was like, 'You know, it's really up to you. You can go beyond the traditional boundaries of linguistics.' I think that's something that I find very rewarding." Upadhye draws inspiration from concepts that are not seen as adjacent to linguistics traditionally, such as behavioral economics and reinforcement learning, and also integrates large language models into her research, something that's uncommon in language production studies when compared to other subfields such as language comprehension.

"Shiva is a great thinker and programmer and experimentalist, who brings a unique mix of creativity and conscientiousness," says language science associate professor Richard Futrell, Upadhye’s advisor. "She’s someone you can count on. I think she’ll go far in our field of science!"

As a small department, language science has a close-knit culture. "Everybody knows everybody really well. The faculty are also very accessible. You can kind of just knock at their doors and talk to them, and it's all very open," says Upadhye. "Most departments don't give you that kind of access."

In addition to the mentorship she's gotten from Futrell, she's also been able to collaborate with Greg Scontras, another associate professor. And Lisa Pearl, professor and founding chair in the department, has given her valuable advice too. "I feel like you can go to them for basically any problem you have or anything you don't understand, any advice that you want, and they'd be very happy to help," says Upadhye.

It was Futrell who instilled a particularly important lesson. "Question everything," she says. "Be skeptical." Nontraditional approaches that go beyond the status quo can be invaluable in research, she says. "If anything seems like it could be relevant, then it's worth exploring."

—Alison Van Houten for UCI Social Sciences