Peter Salovey President | Yale University
Peter Salovey President | Yale University
In a recent interview, Yale University welcomed linguist Simon Charlow to its Faculty of Arts and Sciences as an associate professor of linguistics. Charlow, who has a background in formal semantics and computational linguistics, joins Yale after spending about eight years on the faculty at Rutgers University. He also took a brief detour into private industry with a Boston-based semiconductor company.
Charlow expressed his motivation for this transition: “I wanted to experience how linguistics translated from academia to industry,” he stated. He found that his new colleagues were “phenomenally smart and interested in the same kinds of questions that fascinate me.”
At Yale, Charlow's work involves applying mathematics and computer science to explore how humans encode thoughts into language. He explained his approach: “Most people don’t know what linguists do — it’s not like it’s taught in high school. We study language through a scientific lens.” As a semanticist, Charlow focuses on understanding how complex meanings are constructed from simpler ones using tools from math, logic, and computer science.
Charlow highlighted the intersection between linguistics and computer science, noting that while he is not trained as a computer scientist, he uses concepts from the field to illuminate fundamental questions about language. He sees potential parallels between programming languages created by humans and natural languages that evolve organically.
Discussing recent advancements in artificial intelligence (AI), Charlow addressed the impact of large language models (LLMs) like ChatGPT on linguistic studies. "There is an interesting tension these days between the symbolic models that linguists have used...and the large language models [LLM] that drive ChatGPT," he remarked. While some argue LLMs reduce the need for traditional linguistic methods, Charlow believes symbolic methods remain crucial for understanding language intricacies.
He pointed out that LLMs require significantly more data than young children to learn languages and struggle with aspects of meaning beyond their training scope. Despite these limitations, Charlow suggests LLMs could offer insights into human language learning processes.
Reflecting on his decision to join Yale, Charlow cited the institution's strong faculty in both its linguistics and philosophy departments as key attractions. The department's focus on theoretically informed computational linguistics was particularly appealing to him.
Charlow officially began his role at Yale on July 1, 2024.