By Jackie Jauregui
Artificial Intelligence tools such as ChatGPT form the future of language translation, says Meagan Carter, a Ph.D. candidate in UC Santa Barbara’s Spanish and Portuguese department.
Carter, who has developed a new course called “Translation, Technology, and the Mind,” has found that AI’s results show varying success across specialties such as legal, medical, and community translation, as creative writing proves to be a challenge.
ChatGPT, for instance, knows how to translate a text, but does not understand how to rewrite it to include poetic devices. “AI is big in the field right now. Speech-to-text and media subtitling have been growing and developing very quickly recently,” she said.
She devised the new course so students can stay abreast of these developments. “I want students to understand AI on the basic level,” Carter said. “I want them to understand both its limits and capacities.”
In her sixth year as a Ph.D. candidate, Carter wrote the course to explore the field of language translation and interpreting, both as a profession and as an area of study available to students.
The department regularly offers introductory courses in Hispanic linguistics, translation, and interpreting. But this is the first time a class has been offered that focuses on Carter’s field of research: cognitive translation and interpretation studies — that is to say, what goes on in a person’s mind when they translate a written text or interpret information orally from one language to another.
As part of her dissertation, she aims to understand the relationship between language and trauma, observing a student translator’s cognitive processes and emotional responses when working with traumatic content in court documents that deal with violence and homicide.
This course was offered during the 2023 summer session through UCSB Creative Computing, an interdisciplinary initiative from the Division of Humanities and Fine Arts that offers students technical literacy in emerging technologies across a variety of fields.
“Translation, Technology, and the Mind” offers students a survey of common technologies used by interpreter-translators and lab researchers. That includes the machine translation tools Google Translate and DeepL, an eye tracker and key-logging software. It focuses on the history of digital translation tools, the effectiveness of AI in literary text translation, as well as research methods. The beginning of the week features theory and research, while the latter class is a workshop where students work with tools and technology first hand.
In the third week, local professional Lena Morán-Acereto led a workshop. She is the founder and director of Bridging Voices - Uniendo Voces LLC, a local translation and interpretation provider that offers training workshops for professionals. Morán-Acereto shared her experience as a professional translator-interpreter and evaluated the effectiveness of different technologies.
She explained that it is now a common practice for professional translators to run text through a machine translator first, so that the translator’s role is to edit, to make sure the translation fit a client’s goals precisely. Morán-Acereto has translated natural disaster warnings that run on social media — such as the Thomas Fire in Santa Barbara and Ventura County in 2017 — whereby she had to consider the text character limit on platforms like Twitter.
Carter had been developing the new course since 2019 but didn’t expect just how thoroughly her students would engage with the content. “I was shocked by the dedication of the students,” Carter said. When one student, for instance, wasn’t satisfied with the way the key-logging software Translog II exported its data, they worked to convert it into the coding language R.
Carter says she wants students to understand the processes that happen while an interpreter-translator works, as opposed to merely comparing original texts to their final translated versions. For example, studies show that novices edit as they work, while professionals wait until they’ve finished their draft to make edits, likely because they have more experience.
She wants to give students access to and understanding of tools such as digital glossaries and machine translators, for use later in their careers as translators and interpreters.
With the prospect of re-offering the course, Carter hopes it will grow and develop as technology changes over time. AI chatbots and speech-to-text programs have just recently emerged for professionals, and it remains to be seen whether they will improve efficiency in the practice of translation.
It has also become clear that these new technologies are not equally effective across all languages. Studies show a wider variety and better quality of services in European languages. Though listed under the Spanish and Portuguese department, the course is taught in English and has no language requirements.
“In the future I would like to have a higher reach across different departments and have students of other language backgrounds see [the technologies’] usefulness for other communities,” Carter said.
Jackie Jauregui is a third-year Pre-Political Science and Spanish major at UC Santa Barbara. She is a Web and Social Media intern with the Division of Humanities and Fine Arts.