Emily Bender

Bender recalled that studying French at school and traveling and studying in Europe sparked her interest in languages. At the University of California, Berkeley, linguistics appealed to her more than the French literature she had initially considered. She was particularly drawn to syntax and later encountered different formal grammar theories and computational grammar work at Stanford.⁠[1]

She earned a linguistics bachelor’s degree at Berkeley in 1995 and received its University Medal, followed by master’s and doctoral degrees in linguistics at Stanford. Her doctoral research examined copula absence in African American Vernacular English. Before joining the University of Washington, she worked at Berkeley, Stanford and YY Technologies.⁠[2][3]

Bender joined the University of Washington in 2003 and later became a professor of linguistics. Her work connects multilingual grammar engineering, computational semantics and the social effects of natural language processing.⁠[4]

The University of Washington’s professional master’s program in computational linguistics launched in 2005, with Bender serving as faculty director from its inception. In a 2023 interview, she described combining technical training with attention to real uses and affected communities.⁠[5]

In 2013, she published Linguistic Fundamentals for Natural Language Processing: 100 Essentials from Morphology and Syntax, introducing morphology and syntax for NLP. A second volume, coauthored with Alex Lascarides in 2019, addressed semantics and pragmatics.⁠[6]

Around 2016, she expanded her work on the social impacts of language technology and began teaching an associated ethics course in 2017. She advocates integrating ethical questions throughout NLP training, examining who benefits, who bears costs and whether automation fits the work involved.⁠[4][5]

With Batya Friedman, Bender proposed data statements for NLP in Transactions of the Association for Computational Linguistics. The paper recommends documenting the linguistic and population context of datasets to help assess generalizability and identify risks of systematic bias and exclusion.⁠[7]

In an essay, she traced the principle known as the Bender Rule: researchers should name the languages they study, including English. She argued that avoiding explicit language-specific rules does not establish cross-linguistic applicability, and that evaluation on English alone cannot stand for other languages.⁠[8]

With Alexander Koller, she wrote Climbing towards NLU: On Meaning, Form, and Understanding in the Age of Data. They argued that a system trained only on linguistic form cannot thereby learn meaning, and examined how natural language understanding research should define its training signals and goals. The paper received the ACL 2020 Best Theme Paper award.⁠[9]

On the Dangers of Stochastic Parrots: Can Language Models Be Too Big?, coauthored by Bender, Timnit Gebru, Angelina McMillan-Major and Margaret Mitchell, was presented at ACM FAccT 2021. The authors discussed environmental costs, social biases in training data and the risk that fluent text could lead people to attribute communicative intent to a system.⁠[10]

The University of Washington announced that Bender had been elected a fellow of the American Association for the Advancement of Science, a professional society recognition in her academic career.⁠[11]

In 2024, Bender served as president of the Association for Computational Linguistics and began a listed 2024–2027 term as the University of Washington’s Thomas L. and Margo G. Wyckoff Endowed Professor. She had previously chaired the North American Chapter of the ACL.⁠[2][4]

She and Alex Hanna published The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want. The book criticizes AI hype, examines labor, surveillance and concentrated power, and argues that the public should determine the social goals technology serves. They also cohost a podcast examining AI hype.⁠[12][6]

At the Computer History Museum in Mountain View, California, she debated whether large language models understand language, taking a different position from OpenAI’s Sébastien Bubeck. Eliza Strickland of IEEE moderated. Bender argued against equating a model’s linguistic performance with understanding.⁠[13]

In an interview with Ed Ludlow on Bloomberg Technology, Bender rejected warnings portraying AI as an existential threat to humanity, arguing that doomsday discussions distract from harms already occurring. The program’s description identifies environmental costs, effects on workers, surveillance and misinformation, and emphasizes her call for corporate accountability. These statements represent her position on the risk debate.⁠[14]

Bender’s long-running research includes the LinGO Grammar Matrix, an open-source project providing shared foundations for detailed grammars across languages. She also studies how linguistically annotated materials can support grammar development and extend language technology research beyond English.⁠[4]

Personal website: https://faculty.washington.edu/ebender/ (opens in a new window). University of Washington Department of Linguistics: Official website: https://linguistics.washington.edu/ (opens in a new window). Book website: https://thecon.ai/ (opens in a new window). X: Public account: https://x.com/emilymbender (opens in a new window). LinkedInLinkedInLinkedIn is a social networking platform for professionals. It was founded in 2002 by Reid Hoffman and others, launched in 2003, and listed on the New York Stock Exchange in 2011; after Microsoft acquired it for approximately $26.2 billion in 2016, it became a wholly owned Microsoft subsidiary. Its revenue comes from membership subscriptions, advertising, and recruitment solutions.Open the full entry: Public profile: https://www.linkedin.com/in/ebender/ (opens in a new window).⁠[4][12][15]