Research

I study what neural networks for language understanding know about grammar, primarily using acceptability (grammaticality) judgments.

Recently, I built the Corpus of Linguistic Acceptability (CoLA), a corpus of 10657 examples sentences with acceptability judgments from the linguistics literature. You can access CoLA here: https://nyu-mll.github.io/CoLA/

I also do work in formal semantics and pragmatics and theories of discourse structure. I have worked on the English exclusive particle just and on the QUD theory of relevance.

In undergrad, I worked on Categorial Grammar syntax with Pauline Jacobson to analyze a coordination phenomenon in English known as Right-Node Wrapping.

Papers

“Just” don’t ask: Exclusives and potential questions. In Proceedings of Sinn und Bedeutung 24. 2020. [pdf]

Non-resolving responses to polar questions: A revision to the QUD theory of relevance. With Omar Agha. In Proceedings of Sinn und Bedeutung 24. 2020. [pdf]

Can neural networks acquire a structural bias from raw linguistic data? With Samuel R. Bowman. In Proceedings of CogSci 42. 2020. [pdf] [data]

Are Natural Language Inference Models IMPPRESsive? Learning IMPlicature and PRESupposition. With Paloma Jeretic, Suvrat Bhooshan, and Adina Williams. In Proceedings of ACL. 2020. [pdf] [data]

BLiMP: A Benchmark of Linguistic Minimal Pairs. With Alicia Parrish, Haokun Liu, Anhad Mohananey, Wei Peng, Sheng-Fu Wang, and Samuel R. Bowman. Transactions of the Association for Computational Linguistics (TACL). 2020. [pdf] [data & code]

Investigating BERT’s Knowledge of Language: Five Analysis Methods with NPIs. With Yu Cao, Ioana Grosu, Wei Peng, Hagen Blix, Yining Nie, Anna Alsop, Shikha Bordia, Haokun Liu, Alicia Parrish, Sheng-Fu Wang, Jason Phang, Anhad Mohananey, Phu Mon Htut, Paloma Jeretič, and Samuel R. Bowman. In Proceedings of EMNLP. 2019. [pdf]

Neural Network Acceptability Judgments. With Amanpreet Singh and Samuel R. Bowman. Transactions of the Association for Computational Linguistics (TACL). 2019. [pdf] [data]

Grammatical Analysis of Pretrained Sentence Encoders with Acceptability Judgments. With Samuel R. Bowman. ArXiv preprint. 2019. [pdf] [data]

Verb Argument Structure Alternations in Word and Sentence Embeddings. With Katharina Kann, Adina Williams, Samuel R. Bowman. Proceedings of SCiL 2. 2019. [pdf] [data]

Excluding Imperatives. Unpublished manuscript. 2018. [pdf]

Right-node wrapping: A combinatory account. Proceedings of Workshop “Empirical Advances in Categorial Grammar”, ESSLLI. 2015. [pdf]

Right-Node Wrapping. Undergraduate Thesis. 2015. [pdf]

The Syntax of Coordination and Discontinuity in a Combinatory Categorial Grammar. Presented at Tri-College Undergraduate Linguistics Conference. 2015. [pdf]

Talks

Learning Which Features Matter: RoBERTa Acquires a Preference for Linguistic Generalizations (Eventually). Invited talk at Bar Ilan University NLP Group. October 2020.

Learning grammar from raw data: Case studies with acceptability judgments and neural networks. Invited talk at CLingDing, Indiana University Bloomington. September 2020. [slides]

Can neural networks acquire a structural bias from raw linguistic data? With Samuel R. Bowman. Poster presented at CogSci 42. July 2020.

Are Natural Language Inference Models IMPPRESsive? Learning IMPlicature and PRESupposition. With Paloma Jeretic, Suvrat Bhooshan, and Adina Williams. In Talk given at ACL. July 2020.

BLiMP: The Benchmark of Linguistic Minimal Pairs for English. With Alicia Parrish, Haokun Liu, Anhad Mohananey, Wei Peng, Sheng-Fu Wang, and Samuel R. Bowman. Talk given at the 3rd meeting of SCiL. January 2020.

Discovering Structure: Can neural networks acquire a bias for structural generalizations? With Samuel R. Bowman. Talk given at NYU Syntax Brown Bag. December 2020.

Investigating BERT’s Knowledge of Grammar: Five Analysis Methods with NPIs. Talk given at EMNLP 2019. November 2019. [slides]

Neural Network Acceptability Judgments. Talk given at EMNLP 2019. November 2019. [slides]

Non-Resolving Responses to Polar Questions. With Omar Agha. Poster presented at Sinn und Bedeutung 24. September 2019. [poster]

Just: Don’t Ask the Question. Talk given at Sinn und Bedeutung 24. September 2019. [slides]

Verb Argument Structure Alternations in Word and Sentence Embeddings. With Katharina Kann, Adina Williams, Samuel R. Bowman. Poster presented at SCiL 2. 2019.

Non-Resolving Responses to Polar Questions. With Omar Agha. Talk given at SYNC. 2018.

Just Refuses to Answer the QUD. Talk given at NYU Semantics Group. 2018.

Acceptability Judgments from a Neural Network. With Samuel R. Bowman. Talk given at Annual Meeting of the LSA. 2018.

Right-node wrapping: A combinatory account. Talk given at ESSLLI Workshop “Empirical Advances in Categorial Grammar”. 2015.

The Syntax of Coordination and Discontinuity in a Combinatory Categorial Grammar. Talk given at Tri-College Undergraduate Linguistics Conference. 2015.