There's a Sign for That

Mel Chua grew up deaf. She attended “mainstream” schools for hearing children, learned English and focused solely on spoken language at home. Years passed before she met another deaf person, and it was even longer before she learned to sign.

While studying engineering at Olin College, Chua met Ian Smith, a deaf student who lived in her summer cooperative house. They bonded over their love of computers and “other geeky things” before parting ways to pursue graduate degrees. Years later they met again, but this time they had both learned American Sign Language (ASL).

“We were excited at being able to sign with each other and chatted excitedly in our adopted -- and far easier -- language about what was going on with our lives, our friends and so forth,” Chua said. “And then we started talking about our shared field of electrical/computer engineering and our work in the software realm, and our signing ground to a halt. There we were, two deaf engineers, wanting to sign with each other in ASL and unable to discuss our own profession in that language.”

Their conversation slowly slipped into English.

“As best as I can remember, that was the day we decided -- together -- that this was ridiculous, and that we wanted to work towards being able to sign about engineering with one another,” Chua said.

Chua and Smith are computer science content experts for ASLCore, a project at the National Technical Institute for the Deaf (NTID) designed to fill gaps in the ASL lexicon. The project was created in 2014 by Miriam Lerner, Sarah Cannon and Sarah Schneckenburger, staff interpreters at the Rochester Institute of Technology, of which NTID is one of its nine colleges. The project's website now has new signs for over 1,000 academic words, names and concepts from eight different fields of study.

As content experts, it's up to Chua and Smith to identify computer science terms that don't have ASL translations. At the beginning of an annual two-week boot camp at NTID, they present to the translation team a list of around 150 words that would likely be used in an introductory computer science course.

“We would explain concepts as best we could in the best signing we could muster, with pointers to the English words we leaned on so heavily -- and the translation team would transform our explanations into gorgeous, native-fluent ASL,” Chua said. “We’d go back and forth on connotations, details of usage.”

Chua pushed for translations that were as practical as possible. For example, "I insisted that the sign for ‘sensor’ be one-handed because it would so often need to be signed by an engineer holding a piece of equipment in their other hand," she said.

The same process is repeated for each area of study. One of ASLCore's priorities is ensuring that the Deaf community has full control over the translations. Only the deaf people in the room create the signs and choose which to include on the website.

“It is imperative that this project not be perceived as interpreters taking over ASL, which, unfortunately, has been the case traditionally when they are stuck in the classroom with no resources to guide them,” Lerner said.

Since 2014, individual interpreters, homeschool parents and tutors have picked up on the new signs and begun using them. Lerner and her colleagues have only collected anecdotal feedback so far, but they hope to conduct a study next year to get a better idea of how their signs are used and by how many people.

“The [Rochester Institute of Technology] tutor for the humanities, whose office is close by to mine and who has done this job for at least 15 to 20 years, has said to me several times how much ASLCore helps him do his job better,” said Ruth Anna Spooner, a translator for ASLCore. “He often has students come to him for help with philosophy courses, and now, with the new signs in ASLCore, he is able to explain complicated philosophy concepts much more easily than ever before.”

Each subject area presents different challenges. Philosophy has been particularly difficult for the ASLCore team.

"Taking ideas which are not necessarily grounded in a physical thing and trying to represent them in a 3-D embodied language is quite an undertaking," Lerner said.

Computer science, Chua said, was missing very basic words like “code" and “prototype,” which can be difficult to translate because the terms themselves are seemingly so simple.

“These are just a few examples that might give you an idea of just how fundamental some of these missing signs were,” she said. “I mean, I can’t even teach elementary school electronics without a good sign for ‘resistor.’ Or, for that matter, ‘voltage,’ which previously had simply been signed as the letter ‘V.’” [A sign for resistor existed, but not one many experts favor.]

Spooner added that literature and art history were also in dire need of a greater ASL vocabulary.

“You’ve got names of literary movements and literary time periods, such as Romanticism or Victorian, names of authors such as Dostoyevsky, Twain, Dickens, Austen and so forth, and names of characters in literary works. All of these didn’t have signs before, and many still don’t,” Spooner said. “When discussing one book or one play, we’d have to finger spell the author’s name every time someone mentioned it, and every time someone talked about a character, we’d have to finger spell that character’s name.”

Chua was quick to point out that the assumption that ASL can't capture abstract concepts as well as spoken languages undermines its legitimacy.

“Oftentimes, people say things that are really subtle jabs at the legitimacy and equality of signed languages -- for instance, that it’s somehow innately more difficult to represent an abstract concept in a signed language than a spoken one,” she said. “It’s not -- working out these concepts in a language for the first time is hard, regardless of the language or modality. Think about how long it took Descartes to get to ‘cogito, ergo sum’ or how many pages Hegel wrote about ‘Dasein’ while he was figuring it out.”

ASLCore's process is flexible. Sometimes they revisit past signs and recreate them.

“Once in a while, we will get feedback … that a sign isn’t working,” Spooner said. “Maybe it takes too long to sign, which happened with a few of the philosophy signs, or maybe they felt it didn’t quite fit the concept at hand. When we get a chance, we revisit that sign and make tweaks.”

While most of the feedback is constructive, Lerner said that the project has been criticized by people who assume that the signs are final. She and her team are aware that language is usually built over time, and the ASLCore signs are a much-needed starting point and always open to modification.

“We know that it’s an artificial language ‘dump,’ and by this I mean that natural language growth usually does not entail 1,400 new terms being introduced into a lexicon in four years,” she said.

She and Chua compared the process to the fairly recent development of modern Hebrew.

“In the 1940s when the State of Israel was being established, Hebrew had a lot of words for God, and, as far as I know, no word for ‘compiler,’” Chua said. “They needed to build out the language and develop a practice of talking about computer science in Hebrew -- and they did.”

Part of the project also requires the team to challenge and confront bias against ASL in academia.

“Sometimes people still cling to the belief that English is somehow superior to ASL for teaching academic concepts, and based on this misassumption, they criticize our project,” Spooner said. “It’s a slow process, challenging and changing people’s deep-seated attitudes.”

ASLCore's website includes a dictionary of signs with video demonstrations and descriptions of each entry. It includes technical terms such as verb, neuron, decode, torque and portrait; frequently used names such as René Descartes, Frida Kahlo, Mr. Darcy and Lord Voldemort; and a fittingly complex sign for supercalifragilisticexpialidocious.