Finding the unspoken: Rachelle Boyson

Few people watching the dark-haired undergrad work her laptop amid the noisy ping-pong players inside the Stevenson Recreation Room would realize that groundbreaking research was going on.

But that’s exactly what was happening as Rachelle Boyson, a UC Santa Cruz linguistics major, labored to help solve a communication quirk that has been confounding linguists and computer engineers for years but which any six-year-old can perform without much trouble.

The question she was helping to answer: How do humans understand that which is left unsaid?

How do humans understand that which is left unsaid? That’s the question Rachelle Boyson, a UC Santa Cruz linguistics major, labored to help solve. It's a communication quirk that has been confounding linguists and computer engineers for years but which any six-year-old can perform without much trouble.

Boyson, now 22, came to UC Santa Cruz from Alameda County with a love of language, music, cooking, and poetry. A friend had introduced her to linguistics professor Jim McCloskey and she’d been intrigued by talk of the rigors of a linguistics major at UC Santa Cruz and the idea of examining human communication on a granular level.

“It ended up being a perfect fit,” Boyson said.

She liked the creativity and logic of the discipline, and the way detail was so important. Sometimes, the analysis that was required made her feel like she was figuring out some wonderful puzzle, she says.

By her senior year, Boyson had been hired by McCloskey and associate professor of linguistics Pranav Anand to be part of an undergrad research team helping to create a free, searchable database of an important phenomenon linguists call “sluicing.”

“It’s this big pervasive thing and no one understands it,” said Boyson, who continued to work on the project even after her March 2015 graduation. “I felt like I was doing something that was pioneering.”

Sluicing is what happens when humans leave out part of a who, what, when, where, why and how clause and yet the listener is still able to understand what’s being said.

For instance, someone might say: “She went to the store but I don’t know why.”

What is unsaid, but understood, is the rest of the sentence: but I don’t know why she went to the store.

Or, someone might declare: “We should meet and talk about that” and a listener would reply “when?” Both parties would understand that the single word, “when” actually meant: “when should we meet and talk about that?”

“To take our silence and work out what we mean with our silence?” McCloskey says of this very-human form of communication. “That’s true magic.”

In a world of Siri, Google Now, Amazon Echos, and driverless cars, it’s also an absurdly difficult problem for engineers trying to iron out the bugs of human/computer interaction. And, for scientists, it’s an intriguing puzzle of the human brain.

To that end, McCloskey and Anand won a National Science Foundation grant and began creation of what will be an open, searchable database on sluicing using thousands of articles from the New York Times from the 1990s.

Boyson was one of the database’s first annotators.

Her job was to search articles for examples of sluicing and then break the sentences down into their components. She’d mark antecedents, correlates, predicates, sluices, and interpret the sluice.

“You’d feel like a detective sometimes,” Boyson said.

By the end of summer 2015, Boyson and her coworkers had completed some 4,000 annotations and Boyson had gone on to help co-write a training manual for the seven undergrad researchers who are continuing the annotation project this year.

“The undergrads we have are very smart and very well trained and so are an incredible resource for us,” McCloskey said. “If we had to spend our time doing it (the annotations), we’d never get it finished.” That pool of trained undergrad talent was also part of the reason, he says, the research team won the NSF grant.

For Boyson, the project was a way to put learning into action and to feed her curiosity. She liked the blend of creativity and logic the research entailed.

“It was really cool to take something we all know how to do and utilize forever and look at it in a finer way,” she said.

It was also the kind of hands-on training being sought by tech companies like Google who are doing this type of work in order to improve and create new products, according to McCloskey.

Boyson started a job at Yahoo! In October, doing analysis and annotation. She also still works for McCloskey and Anand on the sluicing project, going back to review annotations done last year and making sure they align with the current annotation guidelines.

“Rachelle is perceptive,” McCloskey says. “She is very careful with matters of detail and she understands things in a deep way. She’s also very well organized.

“Bring those four things together and you have an expert in the kind of work we are doing.”

He says he’s been amazed at the level of talent and intellect UC Santa Cruz students like Boyson possess, at how quickly they work, and the kind of analysis they are able to do.

“For them, it’s an intellectual endeavor,” McCloskey says. “They are at the forefront of an important research project.”